Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
PARALLEL PROCESSING OF A POOL OF DIFFERENT TYPES OF DATA OBJECTS
Document Type and Number:
WIPO Patent Application WO/2020/128590
Kind Code:
A2
Inventors:
SHARMA PRATIK (IN)
Application Number:
PCT/IB2018/060426
Publication Date:
June 25, 2020
Filing Date:
December 20, 2018
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
SHARMA PRATIK (IN)
International Classes:
G06F9/46
Download PDF:
Claims:
Claims

Following is the claim for this invention:-

1. In this invention we have a common shared pool of different types of data objects which have multiple elements and multiple parallel operations that could be performed on them. For each type of data object we compute a task graph which is a Directed Acyclic Graph of tasks that has to be performed in order. The above graph could contain tasks that can be performed in parallel for the corresponding type of data object. The above task graph for different types of data objects is assigned to different executor threads or processes respectively with each executor thread or process pinned to a different processor or running on a different server machine for parallel execution. The common shared pool of different types of data objects maintain a queue or follow the notion of First-Come-First- Serve(FCFS) for each type of data object. Executor process or thread consume the corresponding type of data objects from the common shared pool in First-Come-First-Serve manner. The above novel technique of parallel processing of a pool of different types of data objects is the claim for this invention.

Description:
Parallel Processing Of A Pool Of Different Types Of Data Objects

In this invention we have a common shared pool of different types of data objects which have multiple elements and multiple parallel operations that could be performed on them. For each type of data object we compute a task graph which is a Directed Acyclic Graph of tasks that has to be performed in order. The above graph could contain tasks that can be performed in parallel for the corresponding type of data object. The above task graph for different types of data objects is assigned to different executor threads or processes respectively with each executor thread or process pinned to a different processor or running on a different server machine for parallel execution. The common shared pool of different types of data objects maintain a queue or follow the notion of First-Come-First-Serve(FCFS) for each type of data object. Executor process or thread consume the corresponding type of data objects from the common shared pool in First-Come-First-Serve manner.