Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method and device for improving speculative multithreading

A multi-threading and threading technology, applied in the direction of machine execution devices, concurrent instruction execution, etc., can solve the problems of increasing TLS power consumption, overhead, and reducing TLS performance, so as to avoid thread rollback, reduce power consumption and overhead, and reduce The effect of fallback

Active Publication Date: 2015-11-25
UNIV OF SCI & TECH OF CHINA
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] In view of this, the present invention provides a method and device for improving speculative multi-thread performance, so as to overcome the thread fallback caused by the "immature read" operation in the prior art, resulting in increased power consumption and overhead of TLS, and reduced Problems with TLS performance

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method and device for improving speculative multithreading
  • Method and device for improving speculative multithreading
  • Method and device for improving speculative multithreading

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0060] Please refer to the attached image 3 , the present invention discloses a flow chart of a method for improving speculative multi-thread performance, which mainly includes the following steps:

[0061] Step S101, acquiring the inferred thread message sent by the pioneer thread and the successor thread in the network message.

[0062] Among all the transmitted network messages, the network messages mainly include non-speculative thread messages and speculative thread messages. Generally speaking, the priority of non-speculative thread messages is higher than that of speculative thread messages. Therefore, before performing step S101, it also includes: obtaining Messages for non-speculative threads in network messages and perform delivery.

[0063] After preferably transmitting the non-speculative thread message, execute step S101 to obtain the speculative thread message in the network message, the speculative thread message includes a memory access request message, and t...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an improved thread level speculation (TLS) method and an improved TLS device. The method comprises the following steps of: acquiring a thread speculation message in a network message, and distinguishing and arranging the priorities of access request messages which are sent by a previous thread and a subsequent thread to determine a priority sequence, wherein the priority of the access request message which is sent by the previous thread is higher than that of the access request message which is sent by the subsequent thread; and transmitting the corresponding access request messages to directory nodes according to the determined priority sequence, and reading or writing data at the directory nodes. By adoption of the method provided by the invention, the priorities of access requests in the thread speculation message are distinguished, the priorities of a plurality of threads in a program are distinguished, and the data at the directory nodes can be read or written according to the priority sequence, so squash of the threads is avoided, the squash rate of the threads and the power consumption of TLS are reduced, and the performance of the TLS is stabilized and improved.

Description

technical field [0001] The present invention relates to the field of microprocessor architecture design, and more specifically, relates to a method and device for improving TLS (ThreadLevelSpeculation, speculative multi-thread) performance. Background technique [0002] With the development of semiconductor technology, multiple computing cores can also be placed on a single processor chip, which mainly uses such as figure 1 The design of the on-chip CMP (ChipMultiprocessor, multi-core processor) shown enables multiple threads to run in parallel, thereby continuously improving processor performance. In the process of parallel programming based on CMP, on the one hand, programmers need to develop parallel programs to provide CMP with multiple threads that can run simultaneously; on the other hand, for existing serial programs, they also need to be parallelized . Among them, when parallelizing an existing serial program, the programmer needs to manually create multiple proces...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06F9/38
Inventor 安虹李颀代文博李功明邓博斌吴石磊
Owner UNIV OF SCI & TECH OF CHINA
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products