Zero Overhead Loop Device for SPMD Processing Speed

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Many-core systems face inefficiencies in processing speeds and computational efficiency due to the overhead of auditing loop instructions in Single Program Multiple Data (SPMD) programs, which consume significant processing cycles and resources.

Innovation Solution

The implementation of a Zero Overhead Loop (ZOL) device that performs audits and executes loop body instructions in hardware, eliminating the need for software-based overhead audits and allowing only the loop body to be processed, thereby enhancing processing efficiency and speed.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If software-based loop audits are used in SPMD programs, then processing correctness is maintained, but processing time increases and computational efficiency decreases

Engineering Contradiction:
Improveprocessing correctnessVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent replaces software-based loop audits with a hardware circuit implementation. The hardware circuit performs the audit function through electronic logic operations rather than software instruction execution, thereby maintaining processing correctness while significantly reducing the time required for loop iteration validation in SPMD programs

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces a hardware circuit as an intermediary component between the processor core and the loop execution unit. This hardware intermediary performs the audit function in parallel with the main execution path, eliminating the sequential overhead of software-based audits and improving overall processing efficiency

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If software-based loop audits are used in SPMD programs, then loop iteration validation is performed, but computational resources are consumed

Engineering Contradiction:
Improveloop iteration validationVSAvoidcomputational efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent substitutes software-based validation logic with a dedicated hardware circuit that performs loop iteration validation through electronic comparison and control logic. This hardware implementation maintains accurate validation of loop iterations while freeing up CPU resources for productive computation, thereby improving overall computational efficiency

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The hardware circuit is designed to autonomously perform the audit function without requiring processor intervention. The circuit self-manages the validation logic, comparison operations, and control signal generation, eliminating the need for software-based audits and improving computational efficiency

Inventive Principle:
Principle #25Self-service

3Speed

If hardware-based ZOL device is implemented, then processing speed increases, but device complexity increases

Engineering Contradiction:
Improveprocessing speedVSAvoiddevice complexity
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The patent segments the processing system into distinct functional units: the main processor core and the dedicated hardware ZOL circuit. This segmentation allows the hardware circuit to handle specific loop audit functions independently, improving processing speed for loop operations while containing the added complexity in a modular, dedicated component rather than distributing it throughout the entire system

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS11294690B2Predicated looping on multi-processors for single program multiple data (SPMD) programs
Publication Date: 2022.04.05 INFINEON TECHNOLOGIES AG
  • US11294690B2 patent drawing
  • US11294690B2 patent drawing
  • US11294690B2 patent drawing

AI summary

Single Program, Multiple Data (SPMD) parallel processing of SPMD instructions can be generated among processors assigned to a task in a plurality of threads. The SPMD parallel processing can be increased in speed by performing predicated looping with the SPMD instructions in an activated SPMD mode of operation over a non-SPMD mode. Execution of overhead instructions is removed from the SPMD instructions associated with a thread in order to only execute the loop body of a loop associated with a data element of a data set in an enhanced Zero Loop Overhead (ZOL) device.