NVIDIA GeForce RTX 2080 Ti To GTX 980 Ti TensorFlow Benchmarks With ResNet-50, AlexNet, GoogLeNet, Inception, VGG-16
Lastly are some performance-per-dollar metrics based on the current pricing as obtained from NewEgg. While the GeForce RTX 2080 Ti card is expensive at $1199 USD for the Founder's Edition, the performance-per-dollar doesn't end up being all that bad and largely comes in between the GTX 1080 and GTX 1080 Ti for value with FP16 precision. At FP32, the RTX 2080 Ti tended to be slightly behind the GTX 1080 in value. Delivering the most value of the GPUs I had available for testing tended to be the GeForce GTX 1070 Ti.
If you have NGC setup on your system with Docker and wish to compare your own system's performance to the results in this article, simply install the Phoronix Test Suite and run phoronix-test-suite benchmark 1810070-PTS-NVIDIANG57 for your own fully-automated, side-by-side benchmark comparison. The MONITOR=gpu.temp,sys.power and COST_PERF_PER_DOLLAR= environment variables can be used during the testing process to generate these additional data points. Other interesting Turing benchmarks are on the way.
As always, if you enjoy my daily Linux benchmarking, consider going premium.
If you enjoyed this article consider joining Phoronix Premium to view this site ad-free, multi-page articles on a single page, and other benefits. PayPal or Stripe tips are also graciously accepted. Thanks for your support.