Low power GPGPU computation with imprecise hardware

Massively parallel computation in GPUs significantly boosts performance of compute-intensive applications but creates power and thermal issues that limit further performance scaling. This paper demonstrates significant GPGPU power savings by relaxing application accuracy requirements and enabling th...

Full description

Saved in:
Bibliographic Details
Published in:Proceedings - ACM IEEE Design Automation Conference pp. 1 - 6
Main Authors: Hang Zhang, Putic, Mateja, Lach, John
Format: Conference Proceeding
Language:English
Published: IEEE 01.06.2014
Subjects:
Adders
Approximate Computing
Computational modeling
Floating Point Unit
GPGPU
Graphics processing units
Hardware
Imprecise Hardware
Linear approximation
Measurement
Power demand
Special Function Unit
ISSN:0738-100X
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Massively parallel computation in GPUs significantly boosts performance of compute-intensive applications but creates power and thermal issues that limit further performance scaling. This paper demonstrates significant GPGPU power savings by relaxing application accuracy requirements and enabling the use of low power imprecise hardware (IHW). A synthesized set of novel imprecise floating point arithmetic units is presented. GPGPU-Sim and GPUWattch are used to estimate impacts of IHW units on output quality and system-level power consumption, providing a quality-power tradeoff model for application-specific optimization. Experimental results for a 45 nm process show up to 32% power savings with negligible impacts on output quality.
ISSN:0738-100X
DOI:10.1145/2593069.2593156