Using a dataset of 470 models of graphics processing units (GPUs) released between 2006 and 2021, we find that the amount of floating-point operations/second per $ (hereafter FLOP/s per $) doubles every ~2.5 years. For top GPUs, we find a slower rate of improvement (FLOP/s per $ doubles every 2.95 years), while for models of GPU typically used in ML research, we find a faster rate of improvement (FLOP/s per $ doubles every 2.07 years). GPU price-performance improvements have generally been slightly slower than the 2-year doubling time associated with Moore’s law, much slower than what is implied by Huang’s law, yet considerably faster than was generally found in prior work on trends in GPU price-performance. Our work aims to provide a more precise characterization of GPU price-performance trends based on more or higher-quality data, that is more robust to justifiable changes in the analysis than previous investigations.
Figure 1. Plots of FLOP/s and FLOP/s per dollar for our dataset and relevant trends from the existing literature
Table 1. Summary of our findings on GPU price-performance trends and relevant trends in the existing literature with the 95% confidence intervals in square brackets.
In future work, we intend to build on this work to produce projections of GPU price-performance, and investigate how our findings inform us about the growth in dollar-spending on computing hardware in Machine Learning.
We would like to thank Alyssa Vance, Ashwin Acharya, Jessica Taylor and the Epoch team for helpful feedback and comments.
Trends in GPU price-performance
Link post
Executive Summary
Using a dataset of 470 models of graphics processing units (GPUs) released between 2006 and 2021, we find that the amount of floating-point operations/second per $ (hereafter FLOP/s per $) doubles every ~2.5 years. For top GPUs, we find a slower rate of improvement (FLOP/s per $ doubles every 2.95 years), while for models of GPU typically used in ML research, we find a faster rate of improvement (FLOP/s per $ doubles every 2.07 years). GPU price-performance improvements have generally been slightly slower than the 2-year doubling time associated with Moore’s law, much slower than what is implied by Huang’s law, yet considerably faster than was generally found in prior work on trends in GPU price-performance. Our work aims to provide a more precise characterization of GPU price-performance trends based on more or higher-quality data, that is more robust to justifiable changes in the analysis than previous investigations.
Figure 1. Plots of FLOP/s and FLOP/s per dollar for our dataset and relevant trends from the existing literature
(n=470)
[2.24, 2.72]
[7.45, 9.04]
(n=26)
[1.54, 3.13]
[5.12, 10.39]
(n=57)
[2.54, 3.52]
[8.45, 11.71]
[1.69, 3.62]
[5.60, 12.03]
Table 1. Summary of our findings on GPU price-performance trends and relevant trends in the existing literature with the 95% confidence intervals in square brackets.
In future work, we intend to build on this work to produce projections of GPU price-performance, and investigate how our findings inform us about the growth in dollar-spending on computing hardware in Machine Learning.
We would like to thank Alyssa Vance, Ashwin Acharya, Jessica Taylor and the Epoch team for helpful feedback and comments.