Understanding the Impact of Input Entropy on FPU, CPU, and GPU Power

TL;DR

This paper investigates how input data characteristics significantly influence power consumption in GPU and CPU systems, revealing up to 67% variation, which impacts energy efficiency and system performance.

Contribution

It demonstrates that input data values, their positions, and uniqueness can cause substantial power variations in GPU and CPU workloads, a factor previously underexplored.

Findings

Power variation up to 67% observed on GPU for same algorithm and input size.

Input data properties significantly affect power consumption.

Implications for supercomputer energy efficiency and performance discussed.

Abstract

Power is increasingly becoming a limiting resource in high-performance, GPU-accelerated computing systems. Understanding the range and sources of power variation is essential in setting realistic bounds on rack and system peak power, and developing techniques that minimize energy. While variations arising during manufacturing and other factors like algorithm among others have been previously studied, this work shows that the program inputs can also severely impact the power consumed not only on the GPU but also CPUs. Power variations of up to 67% were observed on an NVIDIA Ampere A100 GPU for the same algorithm (DGEMM benchmark) and input size with different matrix values. Our investigation shows that the values used as matrix elements, their position, and their uniqueness strongly influence power consumption. The implications of this result on supercomputer performance and energy efficiency are further discussed.