NVIDIA GeForce vs. AMD Radeon Vulkan Neural Network Performance With NCNN

Written by Michael Larabel in Display Drivers on 25 September 2020. Page 2 of 2. 20 Comments

The tests were scaling the image to 4x its original size and the benchmarks were with/without TAA.

RealSR-NCNN AMD vs. NVIDIA Vulkan Linux Benchmarks
RealSR-NCNN AMD vs. NVIDIA Vulkan Linux Benchmarks

To much surprise, the Radeon VII was the fastest and even managed to outperform the NVIDIA TITAN RTX with the Vulkan RealSR-NCNN performance! Radeon VII was leading while also being on a fully open-source driver stack. The Radeon RX 5700 series were also performing very strong with the RX 5700 XT being around the RTX 2080 SUPER to RTX 2080 Ti performance level. Very competitive performance out of RADV on Mesa 20.3-devel up against the NVIDIA hardware for this Vulkan workload. Granted, it will be interesting to see how RTX 30 series and RDNA2 stacks up once having the hands on such hardware.

RealSR-NCNN AMD vs. NVIDIA Vulkan Linux Benchmarks

For those wondering about the GPU power consumption under the RealSR-NCNN workload... More data via OpenBenchmarking.org.

While awaiting more tests, those wanting to experiment with RealSR-NCNN can find portable Windows and Linux binaries via GitHub.

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.

Related Articles
About The Author
Michael Larabel

Michael Larabel is the principal author of Phoronix.com and founded the site in 2004 with a focus on enriching the Linux hardware experience. Michael has written more than 20,000 articles covering the state of Linux hardware support, Linux performance, graphics drivers, and other topics. Michael is also the lead developer of the Phoronix Test Suite, Phoromatic, and OpenBenchmarking.org automated benchmarking software. He can be followed via Twitter, LinkedIn, or contacted via MichaelLarabel.com.