SEMICONDUCTOR MEMORIES

BACKGROUND

[001] Semiconductor memories are used in many electronic systems to store data that may be retrieved at a later time. As the demand has increased for electronic systems to be faster, have greater computing ability, and consume less power, semiconductor memories that may be accessed faster, store more data, and use less power have been continually developed to meet the changing needs. Part of the development includes creating new specifications for controlling and accessing semiconductor memories, with the changes in the specifications from one generation to the next directed to improving performance of the memories in the electronic systems.

[002] Semiconductor memories are generally controlled by providing the memories with command signals, address signals, clock signals. The various signals may be provided by a memory controller, for example. The command signals may control the semiconductor memories to perform various memory operations, for example, a read operation to retrieve data from a memory, and a write operation to store data to the memory. The data may be provided between the controller and memories with known timing relative to receipt of an associated command by the memory. The known timing is typically defined by latency information. The latency information may be defined by numbers of clock cycles of system clock signals CK and CKF.

[003] With newly developed memories, the memories may be provided with system clock signals that are used for timing command signals and address signals, for example, and further provided with data clock signals that are used for timing read data provided by the memory and for timing write data provided to the memory. The memories may also provide clock signals to the controller for timing the provision of data provided to the controller.

[004] The timing of various memory commands provided by the controller and received by the memories may be used to control performance of the memories, including the timing of when clock signals are provided, when data is provided, etc. Limitations on the timing of the various memory commands relative to one another may result in less desirable performance by the memories. As such, it may be desirable to have memory commands with flexible timing to provide desirable memory performance.

SUMMARY

[005] Example apparatuses are described. An example apparatus may include a data clock path including an input buffer. The input buffer may be configured to receive a data clock signal when enabled and the data clock path configured to provide an internal clock signal based on the data clock signal. The example apparatus may further include an input/output circuit configured to receive an internal clock signal from the data clock path and provide an access data clock signal based on the internal clock signal, and a command input circuit configured to receive access commands and timing commands associated with the access commands, and further configured to provide internal access commands responsive to receiving the access commands, to provide an internal first timing command responsive to receiving a first timing command of the timing commands, and to provide an internal second timing command responsive to receiving a second timing command of the timing commands. The example apparatus may further include a command decoder coupled to the command input circuit and configured to decode the internal access commands and provide internal access control signals to perform corresponding access operations and further configured to decode the internal first and second timing commands and provide internal timing control signals to enable the input buffer of the data clock path and to control the input/output circuit to provide the access data clock signal. In some examples, each timing command may be associated with a respective access command. In some examples, the first timing command and the second timing command each includes opcodes. In some examples, the opcodes includes a first opcode for a clock synchronization mode and includes a second opcode for an access data clock mode. In some examples, the access command includes a read command. In some examples, the second timing command is limited to immediately preceding the associated access command. In some examples, the first command decoder is configured to provide internal timing control signals to enable the input buffer of the data clock path responsive to the first timing command following the associated access command. In some examples, the data clock path includes a clock divider circuit configured to provide multiphase clock signals based on the data clock signal. In some examples, the input/output circuit includes a clock circuit configured to provide an internal access data clock signal based on the multiphase clock signals. In some examples, the example apparatus may further include a clock path configured to receive a system clock signal and provide internal system clock signals.

[006] Another example apparatus may include a command bus, an address bus, a data bus, a clock bus. a controller configured to provide access commands and timing commands to the command bus, addresses to the address bus, and a data clock signal to the clock bus, and a memory system coupled to the controller through the command, address, data, and clock busses. The memory system may be configured to provide data to the data bus having a timing based on a timing of corresponding access commands and further configured to provide an access data clock signal having a timing based on a timing of the timing commands. A timing command associated with a respective access command may be separated in time from the respective access command by at least one clock cycle of a system clock signal. In some examples, the memory system includes a plurality of memories each coupled to the command, address, data, and clock busses. In some examples, the plurality of memories of the memory system are organized as ranks of memory. In some examples, the example apparatus may further include a plurality of select signal lines. Each select signal line of the plurality of select signal lines may be coupled to a respective one of the plurality of memories of the memory system. In some examples, controller is further configured to provide a first timing command to a first memory of the plurality of memories to enable an input buffer of the first memory, provide a second timing command to a second memory of the plurality of memories to enable an input buffer of the second memory, provide to the first memory a first access command associated with the first timing command, provide an active data clock signal to the first and second memories, and provide to the second memory a second access command associated with the second timing command. The second timing command and the second access command may be separated in time by at least one clock cycle of the system clock signal. In some examples, the first memory is configured to generate a first access data clock signal at the first memory mat is based on the active data clock signal, provide the first access data clock signal, and provide first data from the first memory responsive to the first access command. In some examples, the second memory is configured to generate a second access data clock signal at the second memory that is based on the active data clock signal, provide the second access data clock signal, and provide second data from the second memory responsive to the second access command.

[007] Example methods are described. An example method includes receiving a data clock signal responsive to receiving a timing command, performing an access operation responsive to receiving an access command associated with the timing command, wherein the access command is separated in time from the associated timing command by at least one clock cycle of a system clock signal, and providing an access data clock signal based on the data clock signal. In some examples, the timing command is received before the associated access command. In some examples, the timing command is received after the associated access command. In some examples, the example method may further include enabling an input buffer to receive the data clock signal responsive to the timing command. In some examples, the method may further include providing read data responsive to the access command. Provision of the read data may be synchronized with the access data clock signal. In some examples, the access data clock signal is provided prior to providing the read data. In some examples, the method may further include performing a clock synchronization operation between the system clock signal and the data clock signal responsive to the timing command.

[008] Another example method may include receiving a system clock signal, enabling an input buffer of a first memory responsive to receiving a first timing command, enabling an input buffer of a second memory responsive to receiving a second timing command, receiving at the first memory a first access command associated with the first timing command, and receiving at the first and second memories an active data clock signal, receiving at the second memory a second access command associated with the second timing command. The second timing command and the second access command may be separated in time by at least one clock cycle of the system clock signal. The example method may further include generating a first access data clock signal at the first memory. The first access data clock signal may be based on the active data clock signal. The example method may further include providing the first access data clock signal, providing first data from the first memory responsive to the first access command, and generating a second access data clock signal at the second memory. The second access data clock signal is based on the active data clock signal. The example method may further include providing the second access data clock signal, and providing second data from the second memory responsive to the second access command. In some examples, the first and second timing commands are the same type of timing of command. In some examples, the first and second timing commands are two different types of timing commands. In some examples, the method may further include receiving an active first select signal when receiving the first timing command and the first access command, and receiving an active second select signal when receiving the second timing command and the second access command. In some examples, the second timing command is received at the second memory before the first timing command is received at the first memory. In some examples, the first timing command immediately precedes the first access command. In some examples, the method may further include receiving a third timing command associated with the second access command. A timing of the providing of the second access data clock signal is based on the third timing command. In some examples, the third timing command is received after the second access command. In some examples, the third timing command is separated in time by at least one clock cycle.

[009] Another example method may include providing a timing command to a memory, providing, to the memory, an access command that is associated with the timing command, waiting at least one clock cycle of a system clock signal between providing the timing and access commands to the memory, providing a data clock signal at a time relative to the providing of the timing command, receiving an access data clock signal based on the data clock signal, and receiving data synchronized with the access data clock signal. In some examples, the timing command is provided before the access command. In some examples, the timing command is provided after the access command. In some examples, the access command is a first access command and the timing command is a first timing command, and the example method further including providing a second timing command to a second memory before providing the first access command, and providing a second access command to the second memory. The second access command may be associated with the second timing command. In some examples, the example method may further include providing a third timing command associated with the second timing command. The first timing command and the third timing command may include opcodes enabling a fast synchronization mode and enabling an early access data clock signal mode. BRIEF DESCRIPTION OF THE DRAWINGS

[010] Figure 1 is a block diagram of a system according to an embodiment of the disclosure.

[011] Figure 2 is a block diagram of an apparatus according to an embodiment of the disclosure.

[012] Figure 3 is a block diagram of a clock path and a data clock path according to an embodiment of the disclosure.

[013] Figure 4 is a timing diagram showing a first phase relationship and a second phase relationship between clock signals according to an embodiment of the disclosure.

[014] Figure 5 is a block diagram of a portion of an IO circuit according to an embodiment of the disclosure.

[015] Figures 6A-6D are timing diagrams for various signals during access operations according to an embodiment of the disclosure.

[016] Figures 7A-7D are timing diagrams for various signals during access operations according to an embodiment of the disclosure.

[017] Figures 8 and 9 are timing diagrams showing various signals during access operations for two ranks of memory according to various embodiments of the disclosure.

[018] Figures lOA-1 and 10A-2, I0B, and IOC are timing diagrams showing various signals during access operations for two ranks of memory according to various embodiments of the disclosure.

[019] Figures l lA-1 and 1 1A-2 and l lB-1 and 11B-2 are timing diagrams showing various signals during access operations for two ranks of memory according to various embodiments of the disclosure.

DETAILED DESCRIPTION

[020] Certain details are set forth below to provide a sufficient understanding of examples of the disclosure. However, it will be clear to one having skill in the art that examples of the disclosure may be practiced without these particular details. Moreover, the particular examples of the present disclosure described herein should not be construed to limit the scope of the disclosure to these particular examples. In other instances, well- known circuits, control signals, timing protocols, and software operations have not been shown in detail in order to avoid unnecessarily obscuring the disclosure. Additionally, terms such as "couples" and "coupled" mean that two components may be directly or indirectly electrically coupled. Indirectly coupled may imply that two components are coupled through one or more intermediate components.

[021] Figure 1 is a block diagram of a system 100 according to an embodiment of the disclosure. The system 100 includes a controller 10 and a memory system 105. The memory system 105 includes memories 1 10(0)-110(p) (e.g., "DeviceO" through "Devicep"), where p is a non-zero whole number. The memories 1 10(0)-110(p) are each coupled to the command, address, data, and clock busses. In some embodiments of the disclosure the memories 1 10(0)-1 10(p) are organized as ranks of memory. In such embodiments, the memories may be accessed by the ranks of memory. The controller 10 and the memory system 105 are in communication over several busses. For example, commands and addresses are received by the memory system 105 on a command bus 115 and an address bus 120, respectively, and data is provided between the controller 10 and the memory system 105 over a data bus 125. Various clock signals may be provided between the controller and memory system 105 over a clock bus 130. The clock bus 130 may include signal lines for providing system clock signals CK and CKF received by the memory system 105, data clock signals WCK, and WCKF received by the memory system 105 and an access data clock signal RDQS provided by the memory system 105 to the controller 10. Each of the busses may include one or more signal lines on which signals are provided.

[022] The CK and CKF signals provided by the controller 10 to the memory system 105 are used for timing the provision and receipt of the commands and addresses. The WCK and WCKF signals and the RDQS signal are used for timing the provision of data. The CK and CKF signals are complementary and the WCK and WCKF signals are complementary. Clock signals are complementary when a rising edge of a first clock signal occurs at the same time as a falling edge of a second clock signal, and when a rising edge of the second clock signal occurs at the same time as a falling edge of the first clock signal. The WCK and WCKF signals provided by the controller 10 to the memory system 105 may be synchronized to the CK and CKF signals also provided by the controller 10 to the memory system 105. Additionally, the WCK and WCKF clock signals may have a higher clock frequency than the CK and CKF signals. For example, in some embodiments of the disclosure, the WCK and WCKF signals have a clock frequency that is four times the clock frequency of the CK and CKF signals. The WCK and WCKF signals may be provided by the controller 10 to the memory system 105 continuously during access operations (e.g., WCK always on option enabled) to improve timing performance for the access operations. However, continuously providing the WCK and WCKF signals increases power consumption by the system. Where power consumption may be of concern, the controller 10 does not provide the WCK and WCKF signals continuously (e.g., WCK always on option disabled).

[023] The controller 10 provides commands to the memory system 10S to perform memory operations. Non-limiting examples of memory commands include timing commands for controlling the timing of various operations, access commands for accessing the memory, such as read commands for performing read operations and write commands for performing write operations, mode register write and read commands for performing mode register write and read operations, as well as other commands and operations. The command signals provided by the controller 10 to the memory system 105 further include select signals (e.g., chip select CS signals CSO, CS1, CSp). While all of the memories 110 are provided the commands, addresses, data, and clock signals, the select signals provided on respective select signal lines are used to select which of the memories 110 will respond to the command and perform the corresponding operation. In some embodiments of the disclosure, a respective select signal is provided to each memory 1 10 of the memory system 10S. The controller 10 provides an active select signal to select the corresponding memory 1 10. While the respective select signal is active, the corresponding memory 100 is selected to receive to the commands and addresses provided on the command and address busses 115 and 120.

[024] In operation, when a read command and associated address are provided by the controller 10 to the memory system 10S, the memory 1 10 selected by the select signals receives the read command and associated addresses, and performs a read operation to provide the controller 10 with read data from a memory location corresponding to the corresponding addresses. The read data is provided by the selected memory 110 to the controller 10 according to a timing relative to receipt of the read command. For example, the timing may be based on a read latency (RL) value that indicates the number of clock cycles of the CK and CKF signals (a clock cycle of the CK and CKF signals referenced as tCK) after the read command when the read data is provided by the selected memory 1 10 to the controller 10. The RL value is programmed by the controller 10 in the memories 1 10. For example, the RL value may be programmed in respective mode registers of the memories 110. As known, mode registers included in each of the memories 1 10 may be programmed with information for setting various operating modes and/or select features for operation of the memories. One of the settings may be for the RL value.

[025] In preparation of the selected memory 1 10 providing the read data to die controller 10, the controller provides active WCK and WCKF signals to the memory system 105. The WCK and WCKF signals may be used by the selected memory 110 to generate an access data clock signal RDQS. A clock signal is active when the clock signal transitions between low and high clock levels periodically. Conversely, a clock signal is inactive when the clock signal maintains a constant clock level and does not transition periodically. The RDQS signal is provided by the memory 1 10 performing the read operation to the controller 10 for timing the provision of read data to the controller 10.

[026] The controller 10 may use the RDQS signal for receiving the read data. In some embodiments of the disclosure, the controller 10 has two modes for using the RDQS signal for receiving the read data. In a first mode, the controller 10 may use the RDQS signal to control the timing of circuitry for capturing the read data from the selected memory 1 10. In a second mode, the controller 10 may recover a clock timing from the RDQS signal and generate an internal timing signal based on the recovered timing. The internal timing signal may then be used by the controller 10 to control the timing of circuitry for capturing the read data from the selected memory 1 10.

[027] The controller 10 provides information to the memory system 10S (e.g., in a command) to indicate in which of the modes the RDQS signal will be used by the controller 10. The memory system 105 provides the RDQS signal to the controller 10 with different timing depending on the mode indicated by the controller 10. For example, as will be described in more detail below, the RDQS signal may be provided to the controller 10 with a first timing for the first mode, and provided to the controller 10 with a second timing for the second mode, where the second timing is relatively earlier (e.g., sooner) in comparison to the first mode. The earlier timing of the memory system 105 providing the RDQS signal to the controller 10 may allow greater time for the controller 10 to recover the clock timing from the RDQS signal before the data is provided by the memory system I OS in order to meet the data timing as established by the read latency value RL.

[028] In operation, when a write command and associated address are provided by the controller 10 to the memory system 10S, the memory 1 10 selected by the select signals receives the write command and associated addresses, and performs a write operation to write data from the controller 10 to a memory location corresponding to die corresponding addresses. The write data is provided to the selected memory 110 by the controller 10 according to a timing relative to receipt of the write command. For example, the timing may be based on a write latency (WL) value that indicates the number of clock cycles of the CK and CK.F signals after the write command when the write data is provided to the selected memory 110 by the controller 10. The WL value is programmed by the controller 10 in the memories 1 10. For example, the WL value may be programmed in respective mode registers of the memories 110.

[029] In preparation of the selected memory 1 10 receiving the write data from the controller 10, the controller provides active WCK and WCKF signals to the memory system 105. The WCK and WCKF signals may be used by the selected memory 110 to generate internal clock signals for timing the operation of circuits to receive the write data. The data is provided by the controller 10 and the selected memory 1 10 receives the write data, which is written to memory corresponding to the memory addresses.

[030] Figure 2 is a block diagram of an apparatus according to an embodiment of the disclosure. The apparatus may be a semiconductor device 200, and will be referred as such. In some embodiments, the semiconductor device 200 may include, without limitation, a DRAM device, such as low power DDR (LPDDR) memory integrated into a single semiconductor chip, for example. The semiconductor device 200 may be included in the memory system 205 of Figure 2 in some embodiments of the disclosure. For example, each of the memories 210 may include a semiconductor device 200. The semiconductor device 200 includes a memory die. The die may be mounted on an external substrate, for example, a memory module substrate, a mother board or the like. The semiconductor device 200 may further include a memory array 250. The memory array 250 includes a plurality of banks, each bank including a plurality of word lines WL, a plurality of bit lines BL, and a plurality of memory cells MC arranged at intersections of the plurality of word lines WL and the plurality of bit lines BL. The selection of the word line WL is performed by a row decoder 240 and the selection of the bit line BL is performed by a column decoder 245. Sense amplifiers (SAMP) are located for their corresponding bit lines BL and connected to at least one respective local I/O line pair (LIOT/B), which is in turn coupled to at least respective one main I/O line pair (MIOT/B), via transfer gates (TG), which function as switches.

[031] The semiconductor device 200 may employ a plurality of external terminals that include command and address terminals coupled to a command bus and an address bus to receive command signals COMMAND and address signals ADDRESS, respectively, clock terminals to receive clock signals CK and CKF, data clock terminals to receive data clock signals WCK and WCKF, data terminals DQ, RDQS, DBI, and DMI, power supply terminals VDD, VSS, VDDQ, and VSSQ, and the ZQ calibration terminal (ZQ).

[032] The command terminals and address terminals may be supplied with an address signal and a bank address signal from outside. The address signal and the bank address signal supplied to the address terminals are transferred, via a command/address input circuit 205, to an address decoder 212. The address decoder 212 receives the address signals and supplies a decoded row address signal to the row decoder 240, and a decoded column address signal to the column decoder 245. The address decoder 212 also receives the bank address signal and supplies the bank address signal to the row decoder 240, the column decoder 245.

[033] The command and address terminals may further be supplied with command signals COMMAND from, for example, a memory controller. The command signals COMMAND may be provided as internal command signals ICMD to a command decoder 215 via the command/address input circuit 205. The command decoder 215 includes circuits to decode the internal command signals ICMD to generate various internal signals and commands for performing operations, for example, a row command signal to select a word line and a column command signal to select a bit line. Another example may be providing internal signals to enable circuits for performing operations, such as control signals to enable signal input buffers that receive clock signals. The internal commands also include output and input activation commands, such as a sync command CMDSYNC. [034] When a read command is issued and a row address and a column address are timely supplied with the read command, read data is read from a memory cell in the memory array 2S0 designated by these row address and column address. The read command is received by the command decoder 2 IS, which provides internal commands to input/output circuit 260 so that read data is output to outside from the data terminals DQ, RDQS, DBI, and DMI via read/write amplifiers 2SS and the input/output circuit 260 according to the RDQS clock signals. The read data is provided at a time defined by read latency information RL that may be programmed in the semiconductor device, for example, in a mode register (not shown in Figure 2). The read latency information RL may be defined in terms of clock cycles of the CK clock signal. For example, the read latency information RL may be a number of clock cycles of the CK signal after the read command is received by the semiconductor device 200 when the associated read data is provided.

[035] When the write command is issued and a row address and a column address are timely supplied with this command, then write data is supplied to the data terminals DQ, DBI, and DMI according to the WCK and WCKF clock signals. The write command is received by the command decoder 215, which provides internal commands to the input/output circuit 260 so that the write data is received by data receivers in the input/output circuit 260, and supplied via the input/output circuit 260 and the read/write amplifiers 255 to the memory array 250. The write data is written in the memory cell designated by the row address and the column address. The write data is provided to the data terminals at a time that is defined by write latency WL information. The write latency WL information may be programmed in the semiconductor device 200, for example, in the mode register (not shown in Figure 2). The write latency WL information may be defined in terms of clock cycles of the CK clock signal. For example, the write latency information WL may be a number of clock cycles of the CK signal after the write command is received by the semiconductor device 200 when the associated write data is provided.

[036] Turning to the explanation of the external terminals included in the semiconductor device 200, the clock terminals and data clock terminals are supplied with external clock signals and complementary external clock signals. The external clock signals CK, CKF, WCK, WCKF may be supplied to a clock input circuit 220. When enabled, input buffers included in the clock input circuit 220 receive the external clock signals. For example, an input buffer receives the CK and CKF signals when enabled by a CKE signal from the command decoder 215 and an input buffer receives the WCK and WCKF signals when enabled by a WCKIBEN signal from the command decoder 215. The clock input circuit 220 may receive the external clock signals to generate internal clock signals ICK and IWCK and I WCKF. The internal clock signals ICK and 1WCK and IWCKF are supplied to internal clock circuits 230.

[037] The internal clock circuits 230 includes circuits that provide various phase and frequency controlled internal clock signals based on the received internal clock signals. For example, the internal clock circuits 230 may include a clock path (not shown in Figure 2) that receives the ICK clock signal and provides internal clock signals ICK and ICKD to the command decoder 215. The internal clock circuits 230 may further include a data clock path that receives the IWCK and IWCKF clock signals and provides multiphase clock signals IWCKn based on the internal clock signals IWCK and IWCKF. As will be described in more detail below, the multiphase clock signals IWCKn have relative phases with each other and have a phase relationship with the WCK and WCKF clock signals. The multiphase clock signals IWCKn may also be provided to the input/output circuit 260 for controlling an output timing of read data and the input timing of write data. The input/output circuit 160 may include clock circuits and driver circuits for generating and providing the RDQS signal. The data clock path may also provide a delayed multiphase clock signal IWCKD, which is one of the multiphase clock signals IWCKn further delayed.

[038] A clock synchronization circuit 275 is provided with the delayed multiphase clock signal IWCKD and the sync command CMDSYNC. The clock synchronization circuit 275 provides an output signal SYNCINFO having a logic level that is indicative of a phase relationship between the multiphase clock signals IWCKn and the WCK and WCKF clock signals.

[039] The power supply terminals are supplied with power supply potentials VDD and VSS. These power supply potentials VDD and VSS are supplied to an internal voltage generator circuit 270. The internal voltage generator circuit 270 generates various internal potentials VPP, VOD, VARY, VPERI, and the like and a reference potential ZQVREF based on the power supply potentials VDD and VSS. The internal potential VPP is mainly used in the row decoder 240, the internal potentials VOD and VARY are mainly used in the sense amplifiers included in the memory array 250, and the internal potential VPERI is used in many other circuit blocks. The reference potential ZQVREF is used in the ZQ calibration circuit 26S.

[040] The power supply terminal is also supplied with power supply potential VDDQ.

The power supply potentials VDDQ is supplied to the input/output circuit 260 together with the power supply potential VSS. The power supply potential VDDQ may be the same potential as the power supply potential VDD in an embodiment of the disclosure. The power supply potential VDDQ may be a different potential from the power supply potential VDD in another embodiment of the disclosure. However, the dedicated power supply potential VDDQ is used for the input/output circuit 260 so that power supply noise generated by the input/output circuit 260 does not propagate to the other circuit blocks.

[041] The calibration terminal ZQ is connected to the ZQ calibration circuit 265. The

ZQ calibration circuit 26S performs a calibration operation with reference to an impedance of RZQ, and the reference potential ZQVREF, when activated by the ZQ calibration command ZQ com. An impedance code ZQCODE obtained by the calibration operation is supplied to the input/output circuit 260, and thus an impedance of an output buffer (not shown) included in the input/output circuit 260 is specified.

[042] Figure 3 is a block diagram of a clock path 310 and a data clock path 330 according to an embodiment of the disclosure. The clock path 310 and data clock path 330 may be included in the semiconductor device 300 of Figure 3 in some embodiments of the disclosure. For example the data clock path 330 may be included in the clock input circuit 220 and the internal clock circuit 230 of the semiconductor device 200 of Figure 2. One or both of the clock path 310 and the data clock path 330 may be modified without departing from the scope of the present disclosure.

[043] The clock path 310 may include an input buffer 312 that receives complementary clock signals CK and CKF and provides an internal clock signal ICK. The input buffer 312 may be included in the clock input circuit 220 of Figure 2. The internal clock signal ICK is based on the CK and CKF clock signals. Repeater circuits 314 receive the ICK clock signal and provide an ICK' clock signal to a delay circuit 316. The repeater circuits 314 drive the ICK' clock signal over a clock line from the input buffer 312 to the delay circuit 316. The ICK' clock signal is delayed by the delay circuit 316 to provide a delayed ICK clock signal ICKD. The ICK' and ICKD signals may be used by a command path (not shown) for timing the decoding and provision of internal command signals to perform memory operations (e.g., read, write, etc.).

[044] The data clock path 330 includes an input buffer 352. When enabled by an active enable signal WCKIBEN (e.g., active high logic level), the input buffer 352 receives complementary clock signals WCK and WCKF and provides the complementary internal clock signals IWCK and IWCKF based on the WCK and WCKF clock signals. The receiver circuit 352 may be enabled, for example, by a command decoder responsive to a memory command. In an embodiment of the disclosure, the IWCK and IWCK clock signals have a same clock frequency as a clock frequency of the WCK and WCKF clock signals, and the IWCK clock signal corresponds to the WCK clock signal and the IWCKF clock signal corresponds to the WCKF clock signal. The input buffer 352 may be included in the clock input circuit 220 of Figure 2.

[045] The IWCK and IWCKF clock signals are provided to a clock divider circuit 354 that is configured to provide multiphase clock signals IWCKO, IWCK90, IWCK 180, IWCK270 (collectively referred to as the multiphase clock signals IWCKn). The multiphase clock signals have relative phases to one another, and have a clock frequency that is less than a clock frequency of the WCK and WCKF clock signals (and the IWCK and IWCKF signals). In an embodiment of the disclosure, the IWCKO, IWCK90, IWCK180, and IWCK270 clock signals have a clock frequency that is one-half the clock frequency of the WCK and WCKF clock signals.

[046] In an embodiment of the disclosure, the IWCKO, IWCK90, IWCK 180, and IWCK270 clock signals have a relative phase of 90 degrees to one another. For example, the IWCK90 clock signal has a phase of 90 degrees relative to the IWCKO clock signal, the IWCK 180 clock signal has a phase of 180 degrees relative to the IWCKO clock signal (and a phase of 90 degrees relative to the IWCK90 clock signal), and the IWCK270 clock signal has a phase of 270 degrees relative to the IWCKO clock signal (and a phase of 90 degrees relative to the IWCK 180 clock signal). In such a case, the multiphase clock signals IWCKO, IWCK90, IWCK 180, IWCK270 may be referred to as "quadrature- phase clock signals. [047] The multiphase clock signals are provided to repeater circuits 356. The repeater circuits 356 include a repeater circuit for each of the multiphase clock signals IWCKn. The repeater circuits 356 drive the multiphase clock signals IWCKn over clock lines from the clock divider circuit 354 to a clock distribution circuit 358. The clock distribution circuit 358 provides the multiphase clock signals IWCKn to various circuitries that operate according to the multiphase clock signals. For example, the multiphase clock signals IWCKn may be provided to clock input/output circuits (not shown in Figure 3) to provide and receive data (referenced in Figure 3 as "To DQ block").

[048] As previously described, the IWCKO, IWCK90, IWCK180, IWCK270 signals provided by the clock divider circuit 354 are based on the IWCK and IWCKF signals. The IWCKO, IWCK90, IWCK 180, IWCK270 signals may have a phase relationship relative to the IWCK and IWCKF signals, and likewise, with the WCK and WCKF signals (from which the IWCK and IWCKF signals are based). For example, the multiphase clock signals IWCKO, IWCK90, IWCK 180, and IWCK270 provided by the clock divider circuit 354 may have one of two phase relationships relative to the WCK and WCKF clock signals. A first phase relationship and a second phase relationship are illustrated in Figure 4.

[049] In the first phase relationship, a rising edge 420 of the IWCKO clock signal is associated with a first rising edge 410 of the IWCK clock signal (and the WCK signal, not shown in Figure 4) and a first rising edge of the CK signal, a rising edge 422 of the I WCK90 clock signal is associated with a first falling edge 412 of the IWCK clock signal, a rising edge 424 of the I WCK 180 clock signal is associated with a second rising edge 414 of the IWCK clock signal and a first falling edge of the CK signal, and a rising edge 426 of the IWCK270 clock signal is associated with a second falling edge 416 of the IWCK clock signal. The first phase relationship may be referred to as an "in order" phase relationship.

[050] In the second phase relationship, a falling edge 430 of the IWCKO clock signal is associated with die first rising edge 410 of the IWCK clock signal (and the WCK signal) and a first rising edge of the CK signal, a falling edge 432 of the IWCK90 clock signal is associated with the first falling edge 412 of the IWCK clock signal, a falling edge 434 of the IWCK180 clock signal is associated with the second rising edge 414 of the IWCK clock signal and a first falling edge of the CK signal, and a falling edge 436 of the IWCK270 clock signal is associated with the second falling edge 416 of the IWCK clock signal. The second phase relationship may be referred to as an "out of order" phase relationship.

[051] The first and second phase relationships are maintained even when a clock frequency of the WCK and WCKF (and IWCK and IWCKF) clock signals changes, for example, the clock frequency increases, as shown in Figure 4 following the falling edge 416 of the IWCK clock signal.

[052] The phase relationship of the multiphase clock signals IWCKn provided by the clock divider circuit 2S4 may not be known until a determination is made. The phase relationship of the multiphase clock signals IWCKn may be determined, for example, by evaluating at least one of the multiphase clock signals. The phase relationship may be determined during a WCK-CK synchronization process, which is described in more detail below.

|0S3| Determining the phase relationship of the multiphase clock signals IWCKn to the

WCK and WCKF signals may be needed because proper operation of the semiconductor device 100 may be based on the multiphase clock signals having one of the phase relationships. For example, read data may be provided by the semiconductor device 100 properly when the multiphase clock signals have the "in order" phase relationship. In such an example, when it is determined that the multiphase clock signals IWCKn have the "out of order" phase relationship, various ones of the multiphase clock signals may be switched to provide "in order" multiphase clock signals. As an example, the IWCK 180 clock signal and the IWCK0 clock signal of the out of order multiphase clock signals may be switched and the IWCK270 clock signal and the IWCK90 clock signal of the out of order multiphase clock signals may be switched. As a result, the "out of order" multiphase clock signals are switched into "in order" multiphase clock signals.

[054] Figure 5 is a block diagram of a portion of an IO circuit according to an embodiment of the disclosure. A RDQS clock circuit 510 and a data latch and shift circuit S30 receive multiphase clock signals IWCK0, IWCK90, IWCK 180, and IWCK270 (collectively IWCKn signals). The IWCKn signals may be quadrature clock signals, each clock signal having a 90 degree phase relative to another one of the clock signals (e.g., 0 degree clock signal, 90 degree clock signal, 180 degree clock signal, and 270 degree clock signal). The IWCKn signals may be based on data clock signals WCK and WCKF, and have a clock frequency that is lower than a clock frequency of the WCK and WCKF signals. In some embodiments of the disclosure, the IWCKn signals have half the clock frequency of the WCK and WCKF signals. The multiphase clock signals IWCKn may be provided by a data clock path that receives the WCK signal. For example, in some embodiments of the disclosure, the IWCKn signals may be provided by the data clock path 330 shown in Figure 3.

[055] The RDQS clock circuit 510 provides an internal strobe signal IRDQS based on the IWCKn signals. The IRDQS signal is provided to the driver circuit 520. The driver circuit 520 provides a data strobe signal RDQS based on the IRDQS signal. The RDQS signal may be provided to a device (e.g., a controller 10) for timing the receipt of data by the device. The clock frequency of the RDQS signal may be greater than the clock frequency of the IWCKn signals. In some embodiments of the disclosure, the RDQS signal has a clock frequency that is twice the clock frequency of the IWCKn signals. Where the clock frequency of the IWCKn signals is one-half the clock frequency of the WCK and WCKF signals, the RDQS signal may have the same clock frequency as the WCK and WCKF signals.

[056] In addition to the IWCKn signals, the data latch and shift circuit 530 receives internal data IDO-IDr, where r is a non-zero whole number. The IDO-IDr data may be provided from a memory array. For example, in some embodiments of the disclosure, the ID data is provided from the memory array 250 to an input/output circuit 260 including the data latch and shift circuit 530. The data latch and shift circuit 530 latches and shifts the internal data IDO-IDr based on the IWCKn signals to provide data IDQO- IDQs where s is a non-zero whole number. The IDQO-IDQs data is provided to data driver circuits 540 that drive the IDQO-IDQs data as DQO-DQs data. The data driver circuits 540 may include (s+1) data driver circuits, in particular, one data driver circuit for each of the IDQO-IDQs data.

[057] In operation, the data latch and shift circuit 530 shifts (r+1) bits wide IDO-IDr data into (s+1 ) bits wide IDQO-IDQs data based on the IWCKn signals. The IDQO-IDQs data is then provided by the data driver circuits 540 as (s+1) bits wide DQO-DQs data. The DQO-DQs data may be provided with a timing that corresponds with the RDQS signal. For example, one bit for each of the DQO-DQs data may be provided at rising and falling clock edges of the RDQS signal. As a result, at each edge of the RDQS signal (s+1) bits are output in parallel. In this manner, receipt of (s+1) bits of the DQO-DQs data, for example, by a device, may be timed according to RDQS signal.

[058] As will be described in more detail below, a controller provides a memory system memory commands to access memory (e.g., read or write memory). The memory commands provided for accessing memory include timing command and access commands. As previously described, timing commands may be used to control the timing of various operations, for example, for a corresponding access command. Examples of access commands include a read command and a write command. Examples of timing commands include a CAS command and a MPC command. The timing commands may include opcodes that set various modes of operation during an access operation for an access command. For example, bits of information associated with various opcodes are included in the timing command. An opcode may include one or more bits of the timing command. The opcodes may be identified by a bit position of the timing command. For example, as will be described in more detail below, opcode OP6 of the timing commands may be associated with a RDQS early mode and opcode OP7 may be associated with a WCK.-CK. fast synchronization mode. The respective mode may be enabled by providing a "1" and disabled by providing a "0" for the associated bit included in the timing command.

|0S9] Figures 6-11 are examples of various access operations according to embodiments of the disclosure. The embodiments illustrate the use of timing commands (e.g., CAS command and MPC command) with an access command (e.g., read command). While the embodiments of Figures 6-1 1 are described in the context of read operations, it will be appreciated that the timing commands may be used in the context of write operations without departing from the scope of the disclosure.

[060] Figures 6A-6D are timing diagrams for various signals during access operations according to an embodiment of the disclosure. Figures 6A-6D will be described with reference to read operations for a system including a controller and a memory system. In some embodiments of the disclosure, the system 100 of Figure 1 may be used for the operation described with reference to Figures 6A-6D. Figures 6A-6D will be described with reference to the system 100 of Figure 1 , but the scope of the disclosure is not limited to the particular system 100. The read latency for the read operation of Figures 6A-6D is 12 tCK (e.g., 12 clock cycles of the CK signal). [061] With reference to Figure 6A, at time TaO, a select signal CSO provided by the controller 10 is active to select a memory 1 10 of the memory system 10S that is associated with the CSO signal (e.g., "DeviceO" of the memory system 1 OS). As a result, DeviceO receives a read command READ responsive to a rising clock edge of the CK signal at time TaO. A co mmand/address input circuit of DeviceO receives the READ command and provides it to a command decoder to generate internal control signals to perform a read operation. For example, the command decoder may generate internal control signals to enable a WCK/WCKF input buffer of the DeviceO in preparation for receiving the WCK and WCKF signals from the controller 10. The WCKF signal is not shown in Figures 6A-6D. As previously described, the WCKF signal is complementary to the WCK signal. For the sake of simplicity, the WCK and WCKF signals may be referred to collectively as the WCK signal where applicable for the description of Figures 6A-6D. The WCK signal remains static between times Ta7 and Ta9 (e.g., static period tWCKPREstatic). That is, the WCK signal remains at a known clock level (e.g., at a low clock level) for the time period between times Ta7 and Ta9. At time Ta9, an active WCK signal provided by the controller 10 is received by the DeviceO. The WCK signal may have a first clock frequency followed by a second higher clock frequency (at time TalO), as illustrated in the embodiment of Figure 6A.

[062] Between times Ta9 when the active WCK signal is received by the DeviceO and Tal2 when the DeviceO provides an active access data clock signal RDQS (e.g., time period tWCKPREtoggle), the DeviceO performs WCK-CK synchronization and begins generating internal clock signals based on the WCK signal. For example, internal clock circuits (e.g., a clock divider circuit) may generate multiphase clock signals used for timing internal operations and determine a phase relationship with the WCK signal. The internal clock signals may be used to provide the RDQS signal, such as by a RDQS clock circuit that uses multiphase clock signals based on the WCK signal to generate the RDQS signal. At time Tal 2, the DeviceO provides the active RDQS signal to the controller 10. Also at time Tal 2, or within a time period tWCKDQO, data DQ is provided from the DeviceO by an input/output circuit. The data DQ is provided having a timing synchronized with the RDQS signal. For example, as shown in the embodiment of Figure 6A, a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete (e.g., a 16-bit data burst is shown in Figure 6A). Figure 6A shows the data DQ provided from one data terminal of the DeviceO. While not shown in Figure 6A, data may concurrently be provided from other data terminals of the DeviceO having the same relative timing.

[063] With reference to Figure 6B, at time Ta-1, a select signal CSO provided by the controller 10 is active to select the DeviceO. As a result, the command/address input circuit of DeviceO receives a CAS command for a rising clock edge of the CK signal at time Ta- 1 and receives a read command READ for a rising clock edge of the CK signal at time TaO. The CAS command represents a timing command as previously described. The CAS command immediately precedes an access command (e.g., the READ command), with the CAS command and the associated access command provided as a sequential pair of commands. The CAS command includes opcode OP6=0 to disable a RDQS early mode and OP7=0 to disable a WCK-CK fast synchronization mode. The RDQS early mode and the WCK-CK fast synchronization mode are described in more detail below. The command decoder decodes the CAS and READ commands and generates the internal control signals accordingly. The operation of Figure 6B proceeds similarly to the operation described with reference to Figure 6A.

[064] Following the READ command, the WCK/WCKF input buffer of the DeviceO is enabled in preparation for receiving the WCK and WCKF signals from the controller 10. The WCK signal remains static for the static period tWCKPREstatic between times Ta7 and Ta9. At time Ta9, an active WCK signal provided by the controller 10 is received by the DeviceO, and the DeviceO performs WCK-CK synchronization and generates internal clock signals based on the WCK signal mat are used to provide the RDQS signal. At time Tal2, the DeviceO provides the active RDQS signal to the controller, and provides data DQ within a time period tWCKDQO of time Tal2. As with Figure 6A, the data DQ is provided from the input/output circuit of DeviceO synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete. While Figure 6B shows the data DQ provided from one data terminal of the DeviceO, data may concurrently be provided from other data terminals of the DeviceO having the same relative timing concurrently.

|06S] With reference to Figure 6C, at time Ta-1, a select signal CSO provided by the controller 10 is active to select the DeviceO. As a result, DeviceO receives a CAS command for a rising clock edge of the CK signal at time Ta-1 and receives a read command READ for a rising clock edge of the CK signal at time TaO. The CAS command includes opcode OP6=0 to disable the RDQS early mode and OP7=l to enable the WCK-CK fast synchronization mode. The command decoder decodes the CAS and READ commands and generates the internal control signals to enable the WCK-CK fast synchronization mode and perform the read operation.

[066] The WCK signal may be provided earlier relative to the timing shown in Figures 6A and 6B when the WCK-CK fast synchronization mode is enabled. With the WCK- CK fast synchronization mode enabled, enablement of the WCK/WCKF input buffer of the DeviceO in preparation for receiving the WCK and WCKF signals from the controller 10 begins at time Ta-1 , that is, at the time the CAS command is received by the DeviceO. As shown in Figure 6C, enablement of the WCK/WCKF input buffer occurs over a time period WCKENL between times Ta-1 and Ta2. Starting at time Ta2, the WCK signal remains static (at the low clock level) for the static period tWCKPREstatic between times Ta2 and Ta4. At time Ta4, an active WCK signal provided by the controller 10 is received by the DeviceO, and the DeviceO performs WCK-CK synchronization and generates internal clock signals based on the WCK signal, which may be used to provide the RDQS signal.

[067] In comparison to the WCK timing shown in Figures 6A and 6B where the WCK- CK fast synchronization mode is not enabled, the DeviceO is ready to receive the WCK signal from the controller 10 earlier when the WCK-CK fast synchronization mode is enabled. For example, as shown in the example of Figure 6C, the WCK signal is provided 5 tCKs earlier than for the examples of Figures 6A and 6B. The controller 10 may enable the WCK-CK fast synchronization mode in order to provide the WCK signal earlier to allow the DeviceO to begin generating internal signals based on the WCK signal.

[068] At time Ta 12, the DeviceO provides the active RDQS signal to the controller, and provides data DQ within a time period tWCKDQO of time Tal2. The data DQ is provided from the DeviceO synchronized with the RDQS signal such that a bit of data DQ is provided far each clock edge of the RDQS signal until a data burst is complete. While Figure 6C shows the data DQ provided from one data terminal of the DeviceO, data may also be provided from other data terminals of the DeviceO having the same relative timing concurrently. [069] With reference to Figure 6D, at time Ta-1, a select signal CSO provided by the controller 10 is active to select the DeviceO. As a result, DeviceO receives a CAS command for a rising clock edge of the CK signal at time Ta-1 and receives a read command READ for a rising clock edge of the CK signal at time TaO. The CAS command includes opcode OP6— 1 to enable RDQS early mode and OP7-1 to enable WCK-CK fast synchronization mode. The command decoder decodes the CAS and READ commands and generates the internal control signals to enable the WCK-CK fast synchronization mode and enable the RDQS early mode for the read operation.

[070] The RDQS signal may be provided by the DeviceO earlier relative to the timing shown in Figures 6A-6C when the RDQS early mode is enabled. Additionally, with the WCK-CK last synchronization mode is enabled, the WCK signal may be provided sooner relative to the timing shown in Figures 6A and 6B. With the WCK-CK fast synchronization mode enabled, enablement of the WCK/WCKF input buffer of the DeviceO in preparation for receiving the WCK and WCKF signals from the controller 10 begins at time Ta- 1 , which is the time the CAS command is received by the DeviceO. As shown in Figure 6D, enablement of the WCK/WCKF input buffer occurs over a time period WCKENL between times Ta-1 and Ta2. Starting at time Ta2, the WCK signal remains static for the static period tWCKPREstatic between times Ta2 and Ta4. At time Ta4, an active WCK signal provided by the controller 10 is received by the DeviceO, and during the time period tWCKPREtoggle the DeviceO performs WCK-CK synchronization and generates internal clock signals based on the WCK signal that are used to provide the RDQS signal.

[071] At time Ta6, or within a time period tWCKDQO of time Ta6, the DeviceO provides the active RDQS signal to the controller. In comparison to the RDQS signal timing shown in Figures 6A-6C where the RDQS early mode is not enabled, the RDQS signal is provided earlier when the RDQS early mode is enabled. For example, as shown in the example of Figure 6D, the RDQS signal is provided 5 to 6 tCKs earlier than for the examples of Figures 6A-6C. The controller 10 may enable the RDQS early mode in order to receive the RDQS signal from DeviceO and recover a timing from the RDQS signal and generate an internal timing signal based on the recovered timing. The internal timing signal generated by the controller 10 may be used to time the receipt of the data DQ from the DeviceO. [072] At time Ta 12, the DeviceO provides data DQ within a time period tWCKDQO of time Tal2. The data DQ is provided from the DeviceO synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete. While Figure 6D shows the data DQ provided from one data terminal of the DeviceO, data may also be provided from other data terminals of the DeviceO having the same relative timing concurrently.

[073] In Figures 6A-6D, the time period WCKENL is shown as 3 clock cycles of the WCK signal (3 tCK), the time period tWCKPREstatic is shown as 2 tCK, and the time period tWCKPREtoggle is shown as 3 tCK. Each of the time periods WCKENL, tWCKPREstatic, and tWCKPREtoggle may be the same or different in other embodiments of the disclosure.

[074] Figures 7A-7D are timing diagrams for various signals during access operations according to an embodiment of the disclosure. Figures 7A-7D will be described with reference to read operations for a system including a controller and a memory system. In some embodiments of the disclosure, the system 100 of Figure 1 may be used for the operation described with reference to Figures 7A-7D. Figures 7A-7D will be described with reference to the system 100 of Figure 1 , but the scope of the disclosure is not limited to the particular system 100. The read latency for the read operation of Figures 7A-7D is 9 tCK (e.g., 9 clock cycles of the CK signal).

[075] With reference to Figure 7A, at time TaO, a select signal CSO provided by the controller 10 is active to select a memory of the memory system 105 that is associated with the CSO signal (e.g., "DeviceO" of the memory system 105). As a result, DeviceO receives a read command READ responsive to a rising clock edge of the CK signal at time TaO. A command/address input circuit of DeviceO receives the READ command and provides it to a command decoder to generate internal control signals to perform a read operation. For example, the command decoder may generate internal control signals to enable a WCK/WCKF input buffer of the DeviceO in preparation for receiving the WCK and WCKF signals from the controller 10. The WCKF signal is not shown in Figures 7A-7D. As previously described, the WCKF signal is complementary to the WCK signal. For the sake of simplicity, the WCK and WCKF signals may be referred to collectively as the WCK signal where applicable for the description of Figures 7A- 7D. The WCK signal remains static between times Ta4 and Ta6 (e.g., static period tWCKPREstatic). That is, the WCK signal remains at a known clock level (e.g., at a low clock level) for the time period between times Ta4 and Ta6. At time Ta6, an active WCK signal provided by the controller 10 is received by the DeviceO. The WCK signal may have a first clock frequency followed by a second higher clock frequency (at time Ta7), as illustrated in the embodiment of Figure 7 A.

[076] Between times Ta6 when the active WCK signal is received by the DeviceO and Ta9 when the DeviceO provides an active access data clock signal RDQS (e.g., time period tWCKPREtoggle), the DeviceO performs WCK-CK synchronization and begins generating internal clock signals based on the WCK signal. For example, internal clock circuits (e.g., a clock divider circuit) may generate multiphase clock signals used for timing internal operations and determine a phase relationship with the WCK signal. The internal clock signals may be used to provide the RDQS signal, such as by a RDQS clock circuit that uses multiphase clock signals based on the WCK signal to generate the RDQS signal. At time Ta9, the DeviceO provides the active RDQS signal to the controller 10. Also at time Ta9, or within a time period tWCKDQO, data DQ is provided from the DeviceO by an input/output circuit. The data DQ is provided having a timing synchronized with the RDQS signal. For example, as shown in the embodiment of Figure 7A, a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete (e.g., a 16-bit data burst is shown in Figure 7A). Figure 7A shows the data DQ provided from one data terminal of the DeviceO. While not shown in Figure 7A, data may concurrently be provided from other data terminals of the DeviceO having the same relative timing.

[077] With reference to Figure 7B, at time Ta-3, a select signal CSO provided by the controller 10 is active to select the DeviceO. As a result, the command/address input circuit of DeviceO receives a MPC command for a rising clock edge of the CK signal at time Ta-3. The MPC command represents a timing command as previously described. The MPC command includes opcode OP6=0 to disable a RDQS early mode and OP7=0 to disable a WCK-CK fast synchronization mode. The command decoder decodes the MPC command and generates the internal control signals accordingly. The select signal CSO is active again at time TaO to select the DeviceO. A read command READ provided at time TaO is received by the command/address input circuit of DeviceO for a rising clock edge of the CK signal at time TaO. With the RDQS early mode and the WCK-CK fast synchronization mode disabled, the operation of Figure 7B proceeds similarly to the operation described with reference to Figure 7A.

[078] As shown in Figure 7B, unlike a CAS command, the MPC command is not limited to immediately preceding the READ command. The MPC command in Figure 7B is provided to the DeviceO three tCK before the READ command. As will be described in more detail below, decoupling the MPC command from immediately preceding the READ command may allow for RDQS signal timing that provides controller 10 with sufficient clock cycles of the RDQS signal to recover a timing from the RDQS signal and also satisfy read latency timing for slower CK clock frequencies.

[079] Following the READ command, the WCK/WCKF input buffer of the DeviceO is enabled in preparation for receiving the WCK and WCKF signals from the controller 10. The WCK signal remains static for the static period tWCKPREstatic between times Ta4 and Ta6. At time Ta6, an active WCK signal provided by the controller 10 is received by the DeviceO, and the DeviceO performs WCK-CK synchronization and generates internal clock signals based on the WCK signal that are used to provide the RDQS signal. At time Ta9, the DeviceO provides the active RDQS signal to the controller, and provides data DQ within a time period tWCKDQO of time Ta9. As with Figure 7A, the data DQ is provided from the DeviceO synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete. While Figure 7B shows the data DQ provided from one data terminal of the DeviceO, data may concurrently be provided from other data terminals of the DeviceO having the same relative timing.

[080] With reference to Figure 7C, at time Ta-3, a select signal CS0 provided by the controller 10 is active to select the DeviceO. As a result, DeviceO receives a MPC command for a rising clock edge of the CK signal at time Ta-3. The MPC command includes opcode OP6=0 to disable the RDQS early mode and OP7=l to enable the WCK- CK fast synchronization mode. The command decoder decodes the MPC command and generates the internal control signals to enable the WCK-CK fast synchronization mode. The WCK signal may be provided earlier relative to the timing shown in Figures 7A and 7B when the WCK-CK fast synchronization mode is enabled. The select signal CS0 is active again at time TaO to select the DeviceO. A read command READ provided at time TaO is received by DeviceO for a rising clock edge of the CK signal at time TaO. [081] With the WCK-CK fast synchronization mode enabled, enablement of the WCK/WCKF input buffer by the command decoder of the DeviceO in preparation for receiving the WCK and WCKF signals from the controller 10 begins at time Ta-3, that is, at the time the MPC command is received by the DeviceO. As shown in Figure 7C, enablement of the WCK/WCKF input buffer occurs over a time period WCKENL between times Ta-3 and Ta-1. Starting at time Ta-1, the WCK signal remains static (at the low clock level) for the static period tWCKPREstatic between times Ta-1 and Tal. At time Tal, an active WCK signal provided by the controller 10 is received by the DeviceO, and the DeviceO performs WCK-CK synchronization and generates internal clock signals based on the WCK signal, which may be used to provide the RDQS signal.

[082] In comparison to the WCK timing shown in Figures 7A and 7B where the WCK- CK fast synchronization mode is not enabled, the DeviceO is ready to receive the WCK signal from the controller 10 earlier when the WCK-CK fast synchronization mode is enabled. For example, as shown in the example of Figure 7C, the WCK signal is provided S tCKs earlier than for the examples of Figures 7A and 7B.

[083] At time Ta9, the DeviceO provides the active RDQS signal to the controller, and provides data DQ within a time period tWCKDQO of time Ta9. The data DQ is provided from the DeviceO synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete. While Figure 7C shows the data DQ provided from one data terminal of the DeviceO, data may also be provided from other data terminals of the DeviceO having die same relative timing concurrently.

[084] With reference to Figure 7D, at time Ta-3, a select signal CS0 provided by the controller 10 is active to select the DeviceO. As a result, DeviceO receives a MPC command for a rising clock edge of the CK signal at time Ta-3. The MPC command includes opcode OP6=l to enable RDQS early mode and OP7=l to enable WCK-CK fast synchronization mode. The command decoder decodes the MPC command and generates the internal control signals to enable the WCK-CK fast synchronization mode and enable the RDQS early mode for the access operation.

[085] The RDQS signal may be provided by the DeviceO earlier relative to the timing shown in Figures 7A-7C when the RDQS early mode is enabled. Additionally, with the WCK-CK fast synchronization mode is enabled, the WCK signal may be provided sooner relative to the timing shown in Figures 7A and 7B. The select signal CSO is active again at time TaO to select the DeviceO. A read command READ provided at time TaO is received by DeviceO for a rising clock edge of the CK signal at time TaO.

[086] With the WCK-CK last synchronization mode enabled, enablement of the WCK/WCKF input buffer of the DeviceO in preparation for receiving the WCK and WCKF signals from the controller 10 begins at time Ta-3, which is the time the CAS command is received by the DeviceO. As shown in Figure 7D, enablement of the WCK/WCKF input buffer occurs over a time period WCKENL between times Ta-3 and Ta-1. Starting at time Ta-1, the WCK signal remains static for the static period tWCKPREstatic between times Ta-1 and Tal .

[087] At time Tal, an active WCK signal provided by the controller 10 is received by the DeviceO, and during the time period tWCKPREtoggle the DeviceO performs WCK- CK synchronization and generates internal clock signals based on the WCK signal that are used to provide the RDQS signal. At time Ta3, or within a time period tWCKDQO of time Ta3, the DeviceO provides the active RDQS signal to the controller 10. In comparison to the RDQS signal timing shown in Figures 7A-7C where the RDQS early mode is not enabled, the RDQS signal is provided earlier when the RDQS early mode is enabled. For example, as shown in the example of Figure 7D, the RDQS signal is provided 5 to 6 tCKs earlier than for the examples of Figures 7A-7C. The controller 10 may enable the RDQS early mode in order to receive the RDQS signal from DeviceO and recover a timing from the RDQS signal and generate an internal timing signal based on the recovered timing. The internal timing signal generated by the controller 10 may be used to time the receipt of the data DQ from the DeviceO.

[088] At time Ta9, the DeviceO provides data DQ within a time period tWCKDQO of time Ta9. The data DQ is provided from the DeviceO synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete (e.g., a 16-bit data burst is shown in Figure 7D). While Figure 7D shows the data DQ provided from one data terminal of die DeviceO, data may also be provided from other data terminals of the DeviceO having the same relative timing concurrently.

[089] With reference to timing of the example of Figure 7D and applying the time periods WCKENL, tWCKPREstatic, and tWCKPREtoggle, using a CAS command instead of a MPC command for a read operation would result in the RDQS signal provided by the DeviceO at time Ta6 (instead of at time Ta4 with a MPC command). In such an example, the CAS command would be received by the DeviceO at time Ta- 1 , that is, immediately preceding the READ command at time TaO. With the sum of the time periods WCKENL, tWCKPREstatic, and tWCKPREtoggle being 7 tCK, the earliest the RDQS signal is provided is at time Ta6 (e.g., time Ta-1 plus 7 tCK equals Ta6).

[090] In some systems, the controller 10 may require a minimum number of RDQS clock cycles in order to recover a timing from the RDQS signal and generate an internal timing signal based on the recovered timing. Providing the RDQS signal at time Ta6, which results from using a CAS command for a read operation assuming the timing of the example of Figure 7D, provides 16 clock cycles of the RDQS signal before time (Ta9 + tWCKDQO), which is the time at which the DeviceO begins providing data DQ in order to meet the read latency RL. In contrast, as shown in Figure 7D, the MPC command results in the DeviceO providing the RDQS signal at time Ta4, which provides 24 clock cycles of the RDQS signal before data DQ is provided by the DeviceO at time (Ta9 + tWCKDQO). The additional clock cycles of the RDQS signal prior to providing data DQ may be advantageous for some clock frequencies and when using controllers having a minimum number of RDQS clock cycles for data clock recovery.

[091] In Figures 7A-7D, the time period WCKENL is shown as 2 clock cycles of the

WCK signal (2 tCK), the time period tWCKPREstatic is shown as 2 tCK, and the time period tWCKPREtoggle is shown as 3 tCK. Each of the time periods WCKENL, tWCKPREstatic, and tWCKPREtoggle may be the same or different in other embodiments of the disclosure.

[092] Figures 8 and 9 are timing diagrams showing various signals during access operations for two ranks of memory according to various embodiments of the disclosure. Each rank is represented by a respective device, in particular, RankO corresponds to DeviceO, which is selected by an active select signal CS0 and Rankl corresponds to Device 1 which is selected by an active select signal CS1. In other embodiments of the disclosure, there may be greater than two ranks. Additionally, in some embodiments of the disclosure a rank may include a plurality of devices.

[093] Figures 8 and 9 will be described with reference to read operations for a system including a controller and a memory system. In some embodiments of the disclosure, the system 100 of Figure 1 may be used for the operation described with reference to Figures 8 and 9. Figures 8 and 9 will be described with reference to the system 100 of Figure 1 , but the scope of the disclosure is not limited to the particular system 100. The read latency for the read operation of Figures 8 and 9 is 17 tCK (e.g., 17 clock cycles of the CK signal). The timing diagram of Figure 8 assumes that the WCK always on option is enabled (e.g., WCKaon = 1 for the corresponding mode register setting for the memories 110). With the WCK always on option enabled, the controller 10 provides a continuously active WCK signal after DeviceO and Device 1 are both prepared to receive the WCK signal, as will be described in greater detail below.

[094] With reference to Figure 8, at time Ta-2, a select signal CS0 provided by the controller 10 is active to select the DeviceO (RankO). As a result, a command/address input circuit of DeviceO receives a MPC command for a rising clock edge of the CK signal at time Ta-2. At time Ta-1, a select signal CS1 provided by the controller 10 is active to select the Device 1 (Rankl). As a result, a command/address input circuit of Device 1 receives a MPC command for a rising clock edge of the CK signal at time Ta- I. The MPC commands at times Ta-2 and Ta-1 include opcode OP6=0 to disable RDQS early mode and OP7=l to enable WCK-CK fast synchronization mode. As previously described, the WCK signal may be provided earlier when the WCK-CK fast synchronization mode is enabled in comparison to when the WCK fast synchronization mode is not enabled.

[095] With the WCK-CK fast synchronization mode enabled, the WCK/WCKF input buffers of DeviceO and Device 1 are enabled by receiving the MPC commands in preparation for receiving the WCK and WCKF signals from the controller 10. The WCKF signal is not shown in Figures 8 and 9. As previously described, the WCKF signal is complementary to the WCK signal. For the sake of simplicity, die WCK and WCKF signals may be referred to collectively as the WCK signal where applicable for the description of Figures 8 and 9.

[096] The WCK/WCKF buffer of DeviceO is enabled starting at time Ta-2 and the WCK/WCKF buffer of Device 1 is enabled starting at time Ta-1. Figure 8 illustrates the timing for the Device 1, but not for DeviceO for the sake of simplifying the figure. It will be understood that the timing for enabling the WCK/WCKF buffer of DeviceO is the same as for enabling the WCK/WCKF buffer of Device 1, but beginning and ending 1 tCK earlier than for Device 1. As shown in Figure 8, enablement of the WCK/WCKF input buffer for Device 1 occurs over a time period WCKENL between times Ta-1 and Ta3 (as represented in Figure 8 by the WCK IB enable for Rank 1 becoming active around time Ta3), and enablement of the WCK/WCKF input buffer for DeviceO occurs over a time period WCKENL between times Ta-2 and Ta2 (as represented in Figure 8 by the WCK IB enable for RankO becoming active around time Ta2).

[097] The controller 10 provides a static WCK signal following the latest enabled WCK/WCKF input buffer, which in the example of Figure 8 is the WCK/WCKF input buffer of Devicel . In particular, starting at time Ta3, the WCK signal remains static (at the low clock level) for the static period tWCKPREstatic between times Ta3 and Ta6. At time Ta6, an active WCK signal provided by the controller 10 is received by DeviceO and Devicel. Both DeviceO and Devicel perform WCK-CK synchronization and generate internal clock signals based on the WCK signal that are used to provide the RDQS signal.

[098] Referring back to time TaO, the select signal CSO is active to select the DeviceO so that a read command READ provided at time TaO is received by DeviceO for a rising clock edge of the CK signal. With a read latency of 17 tCKs, data for the READ command at time TaO will be provided by DeviceO following time Tal7. The select signal CS1 is active at time Ta4 to select the Devicel so that a read command READ is received by Devicel for a rising clock edge of the CK signal at time Ta4. With a read latency of 17 tCKs, data for die READ command at time Ta4 will be provided by Devicel following time Ta21.

[099] Following time Tal7 and for the READ command of time TaO (for RankO), the DeviceO provides the active RDQS signal to the controller 10, and provides data DQ within a time period tWCKDQO of time Tal 7. The data DQ is provided from the DeviceO synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete (e.g., a 16-bit data burst is shown in Figure 8). While Figure 8 shows the data DQ provided from one data terminal of the DeviceO, data may also be provided from other data terminals of the DeviceO having the same relative timing concurrently.

[0100] Following time Ta21 and for the READ command of time Ta4 (for Rankl), the Devicel provides the active RDQS signal to the controller 10, and provides data DQ within a time period tWCKDQO of time Ta21. The data DQ is provided from the Devicel synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete (e.g., a 16-bit data burst is shown in Figure 8). While Figure 8 shows the data DQ provided from one data terminal of the Devicel, data may also be provided from other data terminals of the Devicel having the same relative timing concurrently.

[0101] The input buffers of the DeviceO and Devicel remain enabled although no read commands are provided to the DeviceO and Devicel subsequent to the respective read commands. That is, with the WCK always on option enabled in the example of Figure 8, as previously described, the WCK/WCKF input buffers of DeviceO and Devicel remain enabled. However, while not shown in Figure 8, the WCK/WCKF input buffers of DeviceO and Devicel may be disabled using a CAS command or MPC command, with the opcode OP7 = 0, that is, with WCK-CK fast synchronization mode disabled.

[0102] With reference to Figure 9, the timing diagram of Figure 9 assumes that the WCK always on option is disabled (e.g., WCKaon - 0 for the corresponding mode register setting for the memories 110). With the WCK always on option disabled, WCK/WCKF input buffers of DeviceO and Devicel are disabled following completion of a read command. The WCK/WCKF input buffer may remain enabled when another read command is received by the device before completing a previous read command. In contrast, as previously described with reference to Figure 8, when the WCK always on option is enabled, the WCK/WCKF input buffers of DeviceO and Devicel remain enabled and may be disabled when the respective device receives a CAS command or MPC command with opcode OP7 = 0 to disable the WCK-CK fast synchronization mode.

[0103] The timing of the signals shown in Figures 8 and 9 are similar, except for the WCK IB Enable signal for DeviceO and Devicel (RankO and Rankl). For example, following time Tal9, the WCK IB Enable signal for RankO becomes inactive (inactive low logic level) indicating the disablement of the WCK/WCKF input buffer of DeviceO. Similarly, following time Ta23, the WCK IB Enable signal for Rankl becomes inactive (inactive low logic level) indicating the disablement of the WCK/WCKF input buffer of Devicel . The WCK/WCKF input buffers of DeviceO and Devicel are disabled following completion of the respective read commands, as previously described for the WCK always on option disabled (WCKaon=0). However, although not shown in Figure 9, the WCK/WCKF input buffers of DeviceO and Device 1 remain enabled when a read command is received by a device before completion of a previous read command for that device.

[0104] While Figures 8 and 9 show separate MPC commands provided to DeviceO and Device!, in some embodiments of the disclosure, one MPC command may be provided and received by DeviceO and Device 1 simultaneously. In particular, one MPC command may be received by DeviceO and Device! simultaneously by having the select signals CSO and CS1 both active at the time the MPC command is provided. As a result, both DeviceO and Device 1 receive the MPC command simultaneously.

[0105] Figures lOA-1 and 10A-2, 10B, and IOC are timing diagrams showing various signals during access operations for two ranks of memory according to various embodiments of the disclosure. Each rank is represented by a respective device, in particular, RankO corresponds to DeviceO, which is selected by an active select signal CSO and Rankl corresponds to Device 1 which is selected by an active select signal CS1. In other embodiments of the disclosure, there may be greater than two ranks. Additionally, in some embodiments of the disclosure a rank may include a plurality of devices.

[0106] Figures lOA-1 and 10A-2, 10B, and IOC will be described with reference to read operations for a system including a controller and a memory system. In some embodiments of the disclosure, the system 100 of Figure 1 may be used for the operation described with reference to Figures lOA-1 and 10A-2, 10B, and IOC. Figures lOA-l and 10A-2, 10B, and IOC will be described with reference to the system 100 of Figure 1 , but the scope of the disclosure is not limited to the particular system 100. The timing diagrams of Figures lOA-1 and 10A-2, 10B, and IOC assume that the WCK always on option is enabled (e.g., WCKaon = 1 for the corresponding mode register setting). As previously described, with the WCK always on option enabled, the controller 10 provides a continuously active WCK signal after DeviceO and Device 1 are both prepared to receive the WCK signal, as will be described in greater detail below. Additionally, as previously described, with the WCK always on option enabled, the input buffers of DeviceO and Device 1 remain enabled following completion of an access command. As also previously described, the WCK/WCKF input buffers of DeviceO and Device 1 may be disabled using a CAS command or MPC command, with the opcode OP7 - 0, that is, with WCK-CK fast synchronization mode disabled.

[0107] The read latency for the read operations of Figures 10A-1 and 10A-2, 10B, and IOC are different, as will be described in more detail below. The different read latencies for the three read operations results from different clock frequencies of the CK signal (and CKF signal). The clock frequency for the CK signal of Figures lOA-1 and 10A-2 are the fastest (and highest tCK count for the read latency) of the three read operations, and the clock frequency for the CK signal of Figure IOC is the slowest (and lowest tCK count for the read latency) of the three read operations.

[0108] With reference to Figures lOA-1 and 10A-2, the read latency for the read operation is 17 tCK (e.g., 17 clock cycles of the CK signal). Figure lOA-1 is continued on Figure 10A-2 (collectively referred to as Figure 10A). At time Ta-2, a select signal CS1 provided by the controller 10 is active to select Devicel (Rankl). As a result, Device! receives a MPC command for a rising clock edge of the CK signal at time Ta- 2. At time Ta-1, a select signal CSO provided by the controller 10 is active to select the DeviceO (RankO). As a result, DeviceO receives a CAS command for a rising clock edge of the CK signal at time Ta-1. The MPC command at time Ta-2 includes opcode OP6=0 to disable RDQS early mode and OP7=l to enable WCK-CK fast synchronization mode. The CAS command at time Ta-1 includes opcode OP6=l to enable RDQS early mode and OP7=l to enable WCK-CK fast synchronization mode. As previously described, the RDQS signal may be provided by the DeviceO earlier when the RDQS early mode is enabled in comparison to when the RDQS early mode is not enabled. Additionally, the WCK signal may be provided earlier when the WCK-CK fast synchronization mode is enabled in comparison to when the WCK fast synchronization mode is not enabled. The select signal CSO is active at time TaO to select the DeviceO so that a read command READ is received by DeviceO for a rising clock edge of the CK signal at time TaO. With a read latency of 17 tCKs, data for the READ command at time TaO will be provided by DeviceO following time Tal7.

[0109] Figure 10A illustrates the use of the MPC command and the CAS command for access operations. The MPC command at time Ta-2 is provided to set the RDQS early mode and the WCK-CK fast synchronization mode for Devicel. The CAS command at time Ta-1 is used to set the RDQS early mode and the WCK-CK fast synchronization mode for DeviceO, and immediately precedes the READ command at time TaO for DeviceO.

[0110] With the WCK-CK last synchronization mode enabled for both DeviceO and Device 1, the WCK/WCKF input buffers of DeviceO and Device 1 are enabled by receiving the CAS command and the MPC command, respectively, in preparation for receiving the WCK and WCKF signals from the controller 10. The WCKF signal is not shown in Figures lOA-1 and 10A-2, 10B, and IOC. As previously described, the WCKF signal is complementary to the WCK signal. For the sake of simplicity, die WCK and WCKF signals may be referred to collectively as the WCK signal where applicable for the description of Figures lOA-1 and 10A-2, 1 OB, and IOC.

[0111] The WCK/WCKF buffer of Device 1 is enabled starting at time Ta-2 and the WCK/WCKF buffer of DeviceO is enabled starting at time Ta-1. Figure 10A illustrates the timing for the DeviceO, but not for Device 1 for the sake of simplifying the figure. It will be understood mat the timing for enabling the WCK/WCKF buffer of DeviceO is the same as for enabling the WCK/WCKF buffer of DeviceO, but beginning and ending 1 tCK earlier than for DeviceO. As shown in Figure 10A, enablement of the WCK/WCKF input buffer for DeviceO occurs over a time period WCKENL between times Ta-1 and Ta3 (as represented in Figure 10A by the WCK IB enable for RankO becoming active around time Ta3), and enablement of the WCK/WCKF input buffer for Device 1 occurs over a time period WCKENL between times Ta-2 and Ta2 (as represented in Figure 10A by the WCK IB enable for Rankl becoming active around time Ta2).

[0112] The controller 10 provides a static WCK signal following the latest enabled WCK/WCKF input buffer, which in the example of Figure 10A is the WCK/WCKF input buffer of DeviceO. In particular, starting at time Ta3, the WCK signal remains static (at the low clock level) for the static period tWCKPREstatic between times Ta3 and Ta6. At time Ta6, an active WCK signal provided by the controller 10 is received by DeviceO and Device 1. Both DeviceO and Device 1 perform WCK-CK synchronization concurrently and generate internal clock signals based on the WCK signal that are used to provide the RDQS signal. Concurrent WCK-CK synchronization by both DeviceO and Devicel may take less time than performing WCK-CK synchronization sequentially for DeviceO and Devicel . [0113] At time Ta8, or within a time period tWCKDQO of time Ta8, the DeviceO provides the active RDQS signal to the controller 10. As previously described, the RDQS signal may be provided by the DeviceO earlier when the RDQS early mode is enabled in comparison to when the RDQS early mode is not enabled. As previously described, the controller 10 may enable the RDQS early mode in order to receive the RDQS signal from DeviceO and recover a timing from the RDQS signal and generate an internal timing signal based on the recovered timing. The internal timing signal generated by the controller 10 may be used to time the receipt of the data DQ from the DeviceO.

[0114] The select signal CS1 is active at time Ta9 to select the Device 1 so that a read command READ is received by Device 1 for a rising clock edge of the CK signal at time Ta9. With a read latency of 17 tCKs, data for the READ command at time Ta9 will be provided by Device 1 following time Ta26. The select signal CS1 is active again at time Tal 1 to select the Devicel . A MPC command provided at time Tal 1 is received by Device! at a rising clock edge of the CK signal at time Tal 1. The MPC command at time Tal 1 includes opcode OP6-1 to enable RDQS early mode and OP7-1 to enable WCK-CK fast synchronization mode.

[0115] Within a time period tWCKDQO of time Ta8, DeviceO provides the active RDQS signal to the controller 10. The DeviceO also provides data DQ within a time period tWCKDQO of time Tal 7. The data DQ is provided from the DeviceO synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete (e.g., a 16-bit data burst is shown in Figure 10A). While Figure 10A shows the data DQ provided from one data terminal of the DeviceO, data may also be provided from other data terminals of the DeviceO having the same relative timing concurrently.

|0U6] Within a time period tWCKDQO of time Ta20, Devicel provides the active RDQS signal to the controller 10. The Devicel also provides data DQ within a time period tWCKDQO of time Ta26. The data DQ is provided from the Devicel synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete. While Figure 10A shows the data DQ provided from one data terminal of the Devicel , data may also be provided from other data terminals of the Devicel having the same relative timing concurrently. [0117] Figure 10A illustrates the use of the MPC command at time Tal l after an associated READ command at time Ta9. The MPC command may have a timing relative to the associated access command to reduce unnecessary clocking of the RDQS signal provided by Device! . For example, if a CAS command immediately preceding the READ command at time Ta9 is used instead of the MPC command at time Ta 11 , Device 1 would begin providing the RDQS signal at time Tal7 (e.g., Ta8 for the CAS command and 9 tCK of the CK signal (WCKENL + tWCKPREstatic + tWCKPREtoggle) before providing the RDQS signal). However, the RDQS signal for Devicel is not needed until later. Thus, in the present example, using a CAS command and READ command sequential pair instead of a MPC command would result in three tCK of unnecessary RDQS clocking.

[0118] With reference to Figure 10B, the read latency for the read operation is 12 tCK (e.g., 12 clock cycles of the CK signal). At time Ta-2, a select signal CS1 provided by the controller 10 is active to select Devicel (Rankl). As a result, Devicel receives a MPC command for a rising clock edge of the CK signal at time Ta-2. At time Ta-1, a select signal CSO provided by the controller 10 is active to select the DeviceO (RankO). As a result, DeviceO receives a CAS command for a rising clock edge of the CK signal at time Ta- 1. The MPC command at time Ta-2 includes opcode OP6=0 to disable RDQS early mode and OP7=l to enable WCK-CK fast synchronization mode. The CAS command at time Ta-1 includes opcode OP6=l to enable RDQS early mode and OP7=l to enable WCK-CK fast synchronization mode. The select signal CSO is active at time TaO to select the DeviceO so that a read command READ is received by DeviceO for a rising clock edge of the CK signal at time TaO. With a read latency of 12 tCKs, data for the READ command at time TaO will be provided by DeviceO following time Tal2.

[0119] As with the read operation of Figure 10A, Figure 10B illustrates the use of the MPC command and the CAS command for access operations. The MPC command at time Ta-2 is provided to set the RDQS early mode and the WCK-CK fast synchronization mode for Device 1. The CAS command at time Ta- 1 is used to set the RDQS early mode and the WCK-CK fast synchronization mode for DeviceO, and immediately precedes the READ command at time TaO for DeviceO.

[0120] With the WCK-CK fast synchronization mode enabled for both DeviceO and Devicel, the WCK/WCKF input buffers of DeviceO and Devicel are enabled by receiving the CAS command and the MPC command, respectively, in preparation for receiving the WCK and WCKF signals from the controller 10. The WCK/WCKF buffer of Devicel is enabled starting at time Ta-2 and the WCK/WCKF buffer of DeviceO is enabled starting at time Ta-1. As with Figure 10A, Figure 10B illustrates the timing for the DeviceO, but not for Devicel for the sake of simplifying the figure. As shown in Figure 10B, enablement of the WCK/WCKF input buffer for DeviceO occurs over a time period WCKENL between times Ta-1 and Ta2 (as represented in Figure 10B by the WCK IB enable for RankO becoming active around time Ta2), and enablement of die WCK/WCKF input buffer for Devicel occurs over a time period WCKENL between times Ta-2 and Tal (as represented in Figure 1 OB by the WCK IB enable for Rankl becoming active around time Tal).

[0121] The controller 10 provides a static WCK signal following the latest enabled WCK/WCKF input buffer, which in the example of Figure 1 OB is the WCK/WCKF input buffer of DeviceO. In particular, starting at time Ta2, the WCK signal remains static (at the low clock level) for the static period tWCKPREstatic between times Ta2 and Ta4. At time Ta4, an active WCK signal provided by the controller 10 is received by DeviceO and Devicel. Both DeviceO and Devicel perform WCK-CK synchronization concurrently and generate internal clock signals based on the WCK signal that are used to provide the RDQS signal. At time Ta6, or within a time period tWCKDQO of time Ta6, the DeviceO provides the active RDQS signal to the controller 10.

[0122] The select signal CS1 is active at time Ta8 to select the Devicel . A CAS command provided at time Ta8 is received by Devicel at a rising clock edge of the CK signal at time Ta8. The CAS command at time Ta8 includes opcode OP6=l to enable RDQS early mode and OP7-1 to enable WCK-CK fast synchronization mode. The select signal CS1 is again active at time Ta9 to select the Devicel so that a read command READ is received by Devicel for a rising clock edge of the CK signal at time Ta9. With a read latency of 12 tCKs, data for the READ command at time Ta9 will be provided by Devicel following time Ta21.

[0123] In contrast to the read operation of Figure 10A, Figure 10B shows the use of the CAS command and READ command sequential pair for the access operation of Devicel (instead of using an MPC command). Unnecessary clocking of the RDQS signal is avoided in the example of Figure 10B because the read latency is such that the RDQS signal is provided with desirable timing using the CAS command and READ command.

[0124] The DeviceO provides data DQ within a time period tWCKDQO of time Tal2.

The data DQ is provided from the DeviceO synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete (e.g., a 16-bit data burst is shown in Figure 10B). While Figure 10B shows the data DQ provided from one data terminal of the DeviceO, data may also be provided from other data terminals of the DeviceO having the same relative timing concurrently.

[0125] Within a time period tWCKDQO of time TalS, Device 1 provides the active RDQS signal to the controller 10. The Device 1 also provides data DQ within a time period tWCKDQO of time Ta21. The data DQ is provided from the Device 1 synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete. While Figure 10B shows the data DQ provided from one data terminal of the Device 1 , data may also be provided from other data terminals of the Device 1 having the same relative timing concurrently.

[0126] With reference to Figure IOC, the read latency for the read operation is 9 tCK (e.g., 9 clock cycles of the CK signal). At time Ta-4, a select signal CS 1 provided by the controller 10 is active to select Devicel (Rankl) so that Device 1 receives a MPC command for a rising clock edge of the CK signal at time Ta-4. At time Ta-3, a select signal CSO provided by the controller 10 is active to select the DeviceO (RankO) so that DeviceO receives a CAS command for a rising clock edge of the CK signal at time Ta-3. The MPC command at time Ta-4 includes opcode OP6-0 to disable RDQS early mode and OP7=l to enable WCK-CK fast synchronization mode. The CAS command at time Ta-3 includes opcode OP6-1 to enable RDQS early mode and OP7-1 to enable WCK- CK fast synchronization mode. The select signal CSO is active again at time TaO to select the DeviceO so that a read command READ is received by DeviceO for a rising clock edge of the CK signal at time TaO. With a read latency of 9 tCKs, data for the READ command at time TaO will be provided by DeviceO following time Ta9.

[0127] In contrast to the access operations of Figures 10A and 10B, MPC commands are used to begin initialization of the WCK/WCKF input buffers of DeviceO and Devicel to begin concurrent WCK-CK synchronization and the generation of respective RDQS signals. [0128] With the WCK-CK fast synchronization mode enabled for both DeviceO and Devicel, the WCK/WCKF input buffers of DeviceO and Device 1 are enabled by receiving the CAS command and the MPC command, respectively, in preparation for receiving the WCK and WCKF signals from the controller 10. The WCK/WCKF buffer of Devicel is enabled starting at time Ta-2 and the WCK/WCKF buffer of DeviceO is enabled starting at time Ta-1. As with Figures 10A and 10B, Figure IOC illustrates the timing for the DeviceO, but not for Devicel for the sake of simplifying the figure. As shown in Figure IOC, enablement of the WCK/WCKF input buffer for DeviceO occurs over a time period WCKENL between times Ta-3 and Ta-1 (as represented in Figure IOC by the WCK IB enable for RankO becoming active around time Ta-1), and enablement of the WCK/WCKF input buffer for Devicel occurs over a time period WCKENL between times Ta-4 and Ta-2 (as represented in Figure IOC by the WCK IB enable for Rankl becoming active around time Ta-2).

[0129] The controller 10 provides a static WCK signal following the latest enabled WCK/WCKF input buffer, which in the example of Figure IOC is the WCK/WCKF input buffer of DeviceO. In particular, starting at time Ta-1, the WCK signal remains static (at the low clock level) for the static period tWCKPREstatic between times Ta-1 and Tal. At time Tal , an active WCK signal provided by the controller 10 is received by DeviceO and Devicel. Both DeviceO and Devicel perform WCK-CK synchronization concurrently and generate internal clock signals based on the WCK signal that are used to provide the RDQS signal. At time Ta3, or within a time period tWCKDQO of time Ta3, the DeviceO provides the active RDQS signal to the controller 10.

[0130] The select signal CS1 is active at time Ta6 to select the Devicel so that a MPC command provided at time Ta6 is received by Devicel at a rising clock edge of the CK signal at time Ta6. The MPC command at time Ta6 includes opcode OP 6=1 to enable RDQS early mode and OP7= 1 to enable WCK-CK fast synchronization mode. The select signal CS1 is again active at time Ta9 to select the Devicel so that a read command READ is received by Devicel for a rising clock edge of the CK signal at time Ta9. With a read latency of 9 tCKs, data for the READ command at time Ta9 will be provided by Devicel following time Ta9.

[0131] The DeviceO provides data DQ within a time period tWCKDQO of time Ta9.

The data DQ is provided from the DeviceO synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete (e.g., a 16-bit data burst is shown in Figure IOC). While Figure IOC shows the data DQ provided from one data terminal of the DeviceO, data may also be provided from other data terminals of the DeviceO having the same relative timing concurrently.

[0132] Within a time period tWCKDQO of time Tal2, Device 1 provides the active RDQS signal to the controller 10. The Device 1 also provides data DQ within a time period tWCKDQO of time Tal8. The data DQ is provided from the Device 1 synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete. While Figure 10C shows the data DQ provided from one data terminal of the Device 1 , data may also be provided from other data terminals of the Device 1 having the same relative timing concurrently.

[0133] Figures l lA-1 and 1 1A-2 and l lB-1 and 1 1 B-2 are timing diagrams showing various signals during access operations for two ranks of memory according to various embodiments of the disclosure. Each rank is represented by a respective device, in particular, RankO corresponds to DeviceO, which is selected by an active select signal CSO and Rankl corresponds to Device 1 which is selected by an active select signal CS1. In other embodiments of the disclosure, there may be greater than two ranks. Additionally, in some embodiments of the disclosure a rank may include a plurality of devices.

[0134] Figures 1 lA-1 and 1 1A-2 and 1 lB-1 and 1 1B-2 will be described with reference to read operations for a system including a controller and a memory system. In some embodiments of the disclosure, the system 100 of Figure 1 may be used for the operation described with reference to Figures l lA-1 and 11A-2 and l lB-1 and 11B-2. Figures 1 1 A-l and 1 1 A-2 and 1 1 B-I and 1 1 B-2 will be described with reference to the system 100 of Figure 1 , but the scope of the disclosure is not limited to the particular system 100. The timing diagrams of Figures 1 1 A-l and 1 1A-2 and 1 l B-1 and 1 1B-2 assume that the WCK always on option is disabled (e.g., WCKaon = 0 for the corresponding mode register setting). As previously described, with the WCK always on option disabled, the input buffers for the WCK signal of DeviceO and Device 1 are disabled following completion of an access command. [0135] Figures l lA-1 and 11A-2 and l lB-1 and 1 1B-2 illustrate WCK-CK synchronization for RankO and Rankl performed sequentially, in contrast to WCK-CK synchronization for RankO and Rankl performed concurrently.

[0136] With reference to Figure 11 A- 1 and 11 A-2, the read latency for the read operation is 17 tCK (e.g., 17 clock cycles of the CK signal). Figure 1 lA-1 is continued on Figure 11 A-2 (collectively referred to herein as Figure 1 1 A). At time Ta-1, a select signal CSO provided by the controller 10 is active to select DeviceO (RankO). As a result, DeviceO receives a CAS command for a rising clock edge of the CK signal at time Ta-1. The CAS command at time Ta-1 includes opcode OP6=l to enable RDQS early mode and OP7=l to enable WCK-CK fast synchronization mode. The select signal CSO is active at time TaO to select the DeviceO so that a read command READ is received by DeviceO for a rising clock edge of the CK signal at time TaO. With a read latency of 17 tCKs, data for the READ command at time TaO will be provided by DeviceO following time Tal7.

[0137] The WCK/WCKF buffer of DeviceO is enabled starting at time Ta- 1 in response to the CAS command. As shown in Figure 1 1 A, enablement of the WCK/WCKF input buffer for DeviceO occurs over a time period WCKENL between times Ta-1 and Ta3 (as represented in Figure 11 A by the WCK IB enable for RankO becoming active around time Ta3). The controller 10 provides a static WCK signal following the WCK/WCKF input buffer of DeviceO. In particular, starting at time Ta3, the WCK signal remains static (at the low clock level) for the static period tWCKPREstatic between times Ta3 and Ta6. At time Ta6, an active WCK signal provided by the controller 10 is received by DeviceO. DeviceO performs WCK-CK synchronization and generates internal clock signals based on the WCK signal that are used to provide the RDQS signal.

[0138] At time Ta8, or within a time period tWCKDQO of time Ta8, the DeviceO provides the active RDQS signal to the controller 10. As previously described, the RDQS signal may be provided by the DeviceO earlier when the RDQS early mode (opcode OP7=l ) is enabled in comparison to when the RDQS early mode is not enabled. The DeviceO also provides data DQ within a time period tWCKDQO of time Tal 7. The data DQ is provided from the DeviceO synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete (e.g., a 16-bit data burst is shown in Figure 10A). The input buffer for DeviceO is disabled around time Ta20, as represented in Figure 11A by the WCK IB enable for RankO becoming inactive around time Ta20.

[0139] The select signal CS1 is active at time Tal 5 so that a CAS command is received by the Device 1 (Rankl). The CAS command at time TalS includes opcode OP6=l to enable RDQS early mode and OP7-1 to enable WCK-CK fast synchronization mode. The select signal CS1 is active at time Tal 6 to select the Device 1 so that a read command READ is received by Device 1 for a rising clock edge of the CK signal at time Tal 6. With a read latency of 17 tCKs, data for the READ command at time Tal 6 will be provided by Devicel following time Ta33.

[0140] The WCK/WCKF buffer of Device 1 is enabled starting at time Tal S in response to the CAS command. Enablement of the WCK/WCKF input buffer for Devicel occurs over a time period WCKENL between times TalS and Tal9 (as represented in Figure 11 A by the WCK IB enable for Rankl becoming active around time Tal 9). The controller 10 provides a static WCK signal following the WCK/WCKF input buffer of Devicel. In particular, starting at time Tal 9, the WCK signal remains static (at the low clock level) for the static period tWCKPREstatic between times Ta 19 and Ta22. At time Ta22, an active WCK signal provided by the controller 10 is received by Devicel. Devicel performs WCK-CK synchronization and generates internal clock signals based on the WCK signal that are used to provide the RDQS signal.

[0141] At time Ta24, or within a time period tWCKDQO of time Ta24, the Devicel provides the active RDQS signal to the controller 10. The Devicel provides data DQ within a time period tWCKDQO of time Ta33. The data DQ is provided from the Devicel synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete (e.g., a 16-bit data burst is shown in Figure 11 A). The input buffer for Devicel is disabled around time Ta36, as represented in Figure 11A by the WCK IB enable for Rankl becoming inactive around time Ta36.

[0142] In comparison to the timing of Figure 1 1 A, the timing of Figures l lB-1 and 11B- 2 result in data being provided by Devicel (Rankl ) sooner and unnecessary clock cycles of the RDQS signal may be avoided. Figure l lB-1 is continued on Figure 11B-2 (collectively referred to as Figure 1 1 B). As will be described in more detail below, the timing of Figure LIB uses the MPC command, whereas the timing of Figure 1 1A uses the CAS command.

[0143] With reference to Figure 1 1 B, the select signal CSO is active at time TaO to select the DeviceO (RankO) so that a read command READ is received by DeviceO for a rising clock edge of the CK signal at time TaO. With a read latency of 17 tCKs, data for the READ command at time TaO will be provided by DeviceO following time Tal 7. At time Ta2, a select signal CSO provided by the controller 10 is active to select DeviceO. As a result, DeviceO receives a MPC command for a rising clock edge of the CK signal at time Ta2. The MPC command at time Ta-1 includes opcode OP6=l to enable RDQS early mode and OP7=l to enable WCK-CK fast synchronization mode.

[0144] The WCK/WCKF buffer of DeviceO is enabled starting at time Ta2 in response to the MPC command. As shown in Figure 1 IB, enablement of the WCK/WCKF input buffer for DeviceO occurs over a time period WCKENL between times Ta2 and Ta6 (as represented in Figure 1 1 B by the WCK IB enable for RankO becoming active around time Ta6). The controller 10 provides a static WCK signal following the WCK/WCKF input buffer of DeviceO. In particular, starting at time Ta6, the WCK signal remains static (at the low clock level) for the static period tWCKPREstatic between times Ta6 and Ta9. At time Ta9, an active WCK signal provided by the controller 10 is received by DeviceO. DeviceO performs WCK-CK synchronization and generates internal clock signals based on the WCK signal that are used to provide the RDQS signal.

[0145] At time Tal 1, or within a time period tWCKDQO of time Tal 1 , the DeviceO provides the active RDQS signal to the controller 10. The DeviceO also provides data DQ within a time period tWCKDQO of time Tal 7. The data DQ is provided from the DeviceO synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete (e.g., a 16-bit data burst is shown in Figure 10A). The input buffer for DeviceO is disabled around time Ta20, as represented in Figure 11 A by the WCK IB enable for RankO becoming inactive around time Ta20.

[0146] In comparison to the timing of Figure 1 1 A, the number of clock cycles of the RDQS signal in Figure 1 IB is fewer before data is provided by DeviceO. The timing of Figure 1 1 B has 12 fewer clock cycles of the RDQS signal than for the timing of Figure 11 A (e.g., 36 clock cycles versus 24 clock cycles). The fewer clock cycles of the RDQS signal may reduce power consumption where clock cycles in excess of those provided between times Tal 1 and Tal7 are unnecessary for the controller 10 to operate properly.

[0147] The select signal CS1 is active at time Tal 3 to select the Device! so that a read command READ is received by Device 1 for a rising clock edge of the CK signal at time Tal 3. With a read latency of 17 tCKs, data for the READ command at time Tal 3 will be provided by Device 1 following time Ta30. The select signal CS1 is active again at time TalS so that a MPC command is received by the Device 1 (Rankl). The MPC command at time TalS includes opcode OP6=l to enable RDQS early mode and OP7=l to enable WCK-CK fast synchronization mode.

[0148] The WCK/WCKF buffer of Device 1 is enabled starting at time Tal 5 in response to the CAS command. Enablement of the WCK/WCKF input buffer for Device! occurs over a time period WCKENL between times Tal 5 and Tal9 (as represented in Figure 11 B by the WCK IB enable for Rankl becoming active around time Tal 9). The controller 10 provides a static WCK signal following the WCK/WCKF input buffer of Devicel. In particular, starting at time Tal 9, the WCK signal remains static (at the low clock level) for the static period tWCKPREstatic between times Tal 9 and Ta22. At time Ta22, an active WCK signal provided by the controller 10 is received by Devicel. Devicel performs WCK-CK synchronization and generates internal clock signals based on the WCK signal that are used to provide the RDQS signal.

[0149] At time Ta24, or within a time period tWCKDQO of time Ta24, the Devicel provides the active RDQS signal to the controller 10. The Devicel provides data DQ within a time period tWCKDQO of time Ta30. The data DQ is provided from the Devicel synchronized with the RDQS signal such that a bit of data DQ is provided for each clock edge of the RDQS signal until a data burst is complete (e.g., a 16-bit data burst is shown in Figure 1 IB). The input buffer for Devicel is disabled around time Ta33, as represented in Figure 1 1 B by the WCK IB enable for Rankl becoming inactive around time Ta33.

[0150] Significantly, the READ command is received by Devicel earlier for the timing of Figure 1 IB than for the timing of Figure 1 1 A (e.g., time Tal 3 versus time Tal 6). As a result, data may be provided by Devicel earlier for the timing of Figure 1 IB than for the timing of Figure 1 1A (e.g., time Ta30 versus time Ta33). Additionally, the MPC command at time Tal 5 results in a reduced number of clock cycles of the RDQS signal before data is provided by Device 1. The READ command may be received by Device 1 earlier because the MPC command is not limited to immediately preceding an associated access command, as is the case for a CAS command. The MPC command may be received at a time before or after the associated READ command to enable the input buffer of Device 1 so that unnecessary clock cycles of the RDQS signal may be avoided. |01S1| Thus, as illustrated by Figures 11A and 1 IB, by using the MPC command, Devicel may provide data sooner than when using a CAS command (e.g., Figure 1 1 A) and unnecessary clock cycles of the RDQS signal may be avoided by timing the MPC command accordingly.

[0152] While the previous embodiments of Figures 6-1 1 have been described in the context of read operations, embodiments of the disclosure may also be applied in the context of other memory access operations. For example, the use of the MPC and CAS commands may be used for write operations. Rather than receiving read commands from the controller and providing data to the controller, DeviceO and Devicel receive write commands from the controller and receive data from the controller to be stored in memory.

[0153] Figures 6-11 illustrate the flexibility provided by using MPC commands to perform access operations, for example, including single rank access operations and rank-to-rank access operations, to accommodate different clock frequencies of the CK signal. Unlike the CAS command that immediately precedes an associated access command (e.g., READ command, WRITE command, etc.), the MPC command may be provided and received at times separated from (e.g., not immediately preceding or following) an associated access command. As previously illustrated and described, the MPC command may precede an associated access command or may follow an associated access command, and may be separated in time from the associated access command by at least one clock cycle of a system clock signal (e.g., the CK signal). However, the MPC command may also immediately precede an associated access command or may also immediately follow an associated access command. As a result, the MPC command may be used to provide flexible timing.

|01S4| From the foregoing it will be appreciated that, although specific embodiments of the disclosure have been described herein for purposes of illustration, various modifications may be made without deviating from the spirit and scope of the disclosure. Accordingly, the scope disclosure should not be limited any of the specific embodiments described herein.