Incredible Linux GPU Compute Performance Review Offers NVIDIA GeForce RTX 3080

Yesterday I lastly obtained a GeForce RTX 3080 graphics card from NVIDIA for having the ability to ship the primary of our Linux benchmarks on the brand new RTX 30 Ampere collection. What is instantly clear is the massive efficiency uplift for OpenCL and CUDA workloads with the RTX 3080 in comparison with its predecessors. The uncooked efficiency and even performance-per-dollar is staggering out of the GeForce RTX 3080 with the preliminary checks carried out on Ubuntu Linux. Linux gaming benchmarks shall be out within the days forward however for now could be a take a look at the RTX 3080 compute efficiency throughout dozens of benchmarks and going way back to the GeForce GTX 980 collection for comparability.

For these not paying consideration in latest weeks, the GeForce RTX 3080 is now accessible as NVIDIA’s new providing coming in on the $699 USD value level. The GeForce RTX 3080 options 8,704 CUDA cores with a 1.44GHz base clock and 1.71GHz enhance clock. The GeForce RTX 3080 has 10GB of GDDR6X reminiscence on a 320-bit interface. Ampere brings huge enhancements to its ray-tracing and tensor cores whereas additionally now that includes PCI Specific 4.zero connectivity, improved video encode/decode capabilities together with AV1 decode, and an assortment of different architectural developments with Ampere.

The GeForce RTX 3080 contains a new 12-pin energy connector or for current energy provides an adapter is included for two x 8-pin PCI Specific assist. The RTX 3080 is sort of energy hungry with a rated graphics card energy of 320 Watts.

Yesterday the RTX 3080 arrived courtesy of NVIDIA and was instantly loaded with Linux checks. Sadly we don’t but have an RTX 3090 or any ETA on after we might have that higher-tier card, however the RTX 3080 efficiency already is mighty spectacular. At the least within the weeks of ready for a GeForce RTX 3080 assessment pattern, I’ve already accomplished my re-testing of prior graphics playing cards and thus have already got the preliminary batch of compute numbers to share at this time. As mentioned earlier, the Linux gaming benchmarks shall be out later this week in addition to different RTX 30 Linux benchmarks to observe within the weeks forward.

Because of many CUDA/NVIDIA API targeted benchmarks, on this article is just a take a look at the NVIDIA GPU compute efficiency with none Radeon metrics — clearly within the forthcoming Linux gaming outcomes would be the related Radeon figures. Significantly with Radeon Open eCosystem (ROCm) nonetheless not taking part in correctly on the latest-generation Navi graphics playing cards, even for the OpenCL benchmarks on this article it wasn’t attainable to offer the comparability factors on the RX 5700 collection. In any case as soon as the ROCm assist is in higher form for Navi/RDNA shall be such a compute comparability.

With the time ready for the RTX 3080 additionally allowed re-testing NVIDIA GPUs going again to Maxwell. For at this time’s testing the graphics playing cards benchmarked included:

– GTX 980
– GTX 980 Ti
– GTX TITAN X
– GTX 1060
– GTX 1070
– GTX 1070 Ti
– GTX 1080
– GTX 1650
– GTX 1650 SUPER
– GTX 1660
– GTX 1660 SUPER
– GTX 1660 Ti
– RTX 2060
– RTX 2060 SUPER
– RTX 2070
– RTX 2070 SUPER
– RTX 2080
– RTX 2080 SUPER
– RTX 2080 Ti
– TITAN RTX
– RTX 3080

The check system used was powered by an AMD Ryzen 9 3950X with Asus ROG CROSSHAIR VIII HERO motherboard, 16GB of DDR4-3600 Corsair system reminiscence, 2TB Corsair Power MP600 NVMe storage, and the varied graphics playing cards below check. On the Software program facet was Ubuntu 20.04.1 LTS with the Linux 5.Four kernel and P-State efficiency CPU frequency scaling governor.

The GeForce RTX 3080 was fortunately operating together with the NVIDIA 455.23.05 beta Linux graphics driver and CUDA 11.1. Throughout these GPU compute checks at the least the present NVIDIA Linux driver was operating properly with the GeForce RTX 3080 and didn’t encounter any Linux GPU driver points. In fact, any driver points usually tend to come up throughout our gaming checks within the days forward to remain tuned.

Examined workloads for this text through the Phoronix Take a look at Suite included ArrayFire, Caffe, cl-mem, clpeak, FAHBench, FinanceBench, Geekbench, GROMACS, Leela Chess Zero, LuxCoreRender, Mixbench, NAMD, NCNN, PlaidML, ReakSR-NCNN, RedShift, and VkFFT. The Phoronix Take a look at Suite was additionally monitoring the GPU energy consumption and GPU core temperature throughout testing on per-test foundation in addition to in flip producing performance-per-Watt metrics.