Title:
COMPUTE-INTENSIVE KERNEL GENERATOR, MICRO-KERNEL CODE CACHE, FUSED KERNEL GENERATOR AND CYCLIC DEPENDENCE FREE GRAPH PARTITIONING FOR DEEP LEARNING WORKLOADS
Document Type and Number:
WIPO Patent Application WO/2023/108894
Kind Code:
A1
Abstract:
Systems, apparatuses and methods may provide for technology that identifies a data layout associated with input tensors and output tensors, generates a micro-kernel based at least in part on the data layout, and generates a nested outer loop for a kernel, wherein the micro-kernel performs one or more subtasks associated with a task represented by the kernel. The technology also includes micro-kernel code caches, fused kernel generators and cyclic dependence free graph partitioning for deep learning workloads.
More Like This:
| JP2005513612 | Processor architecture that selectively uses finite state machine control code |
| JPS53107250 | MICROPROGRAM CONTROL SYSTEM |
| JPH08263284 | DIGITAL COMPUTER SYSTEM |
Inventors:
LI JIANHUI (US)
QIN ZHENNAN (CN)
GONG JIONG (CN)
CUI JINGZE (CN)
MEI YIJIE (CN)
SONG YUNFEI (CN)
QIN ZHENNAN (CN)
GONG JIONG (CN)
CUI JINGZE (CN)
MEI YIJIE (CN)
SONG YUNFEI (CN)
Application Number:
PCT/CN2022/077751
Publication Date:
June 22, 2023
Filing Date:
February 24, 2022
Export Citation:
Assignee:
INTEL CORP (US)
LI JIANHUI (US)
QIN ZHENNAN (CN)
GONG JIONG (CN)
CUI JINGZE (CN)
MEI YIJIE (CN)
SONG YUNFEI (CN)
LI JIANHUI (US)
QIN ZHENNAN (CN)
GONG JIONG (CN)
CUI JINGZE (CN)
MEI YIJIE (CN)
SONG YUNFEI (CN)
International Classes:
G06F9/22; G06F8/41; G06F9/38; G06F17/16
Domestic Patent References:
| WO2016054303A1 | 2016-04-07 |
Foreign References:
| US20200410318A1 | 2020-12-31 | |||
| US20210049231A1 | 2021-02-18 | |||
| US20210117806A1 | 2021-04-22 |
Attorney, Agent or Firm:
BEIJING EAST IP LTD. (CN)
Download PDF: