GPU compute

qemu testing on Ubuntu 20.04 via the Phoronix Test Suite.

HTML result view exported from: https://openbenchmarking.org/result/2011263-FI-GPUCOMPUT67.

GPU computeProcessorMotherboardChipsetMemoryDiskGraphicsAudioNetworkOSKernelOpenCLVulkanCompilerFile-SystemSystem LayerGTX 960 ovmfIntel Core i7-3770 (1 Core)QEMU Standard PC (Q35 + ICH9 2009) (0.0.0 BIOS)Intel 82G33/G31/P35/P31 + ICH92048MB69GB QEMU HDDNVIDIA GeForce GTX 960 2GBIntel 82801IRed Hat Virtio deviceUbuntu 20.045.4.0-48-generic (x86_64)OpenCL 1.2 CUDA 11.0.2281.2.133GCC 9.3.0ext4qemuOpenBenchmarking.org- CPU Microcode: 0x1- itlb_multihit: Not affected + l1tf: Mitigation of PTE Inversion + mds: Vulnerable; SMT Host state unknown + meltdown: Vulnerable + spec_store_bypass: Vulnerable + spectre_v1: Vulnerable: __user pointer sanitization and usercopy barriers only; no swapgs barriers + spectre_v2: Vulnerable STIBP: disabled + srbds: Unknown: Dependent on hypervisor status + tsx_async_abort: Not affected

GPU computeblender: BMW27 - CUDAblender: BMW27 - NVIDIA OptiXGTX 960 ovmf285.03335.86OpenBenchmarking.org

Blender

Blend File: BMW27 - Compute: CUDA

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: BMW27 - Compute: CUDAGTX 960 ovmf60120180240300SE +/- 3.58, N = 3285.03

Blender

Blend File: BMW27 - Compute: NVIDIA OptiX

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: BMW27 - Compute: NVIDIA OptiXGTX 960 ovmf70140210280350SE +/- 12.51, N = 9335.86

Phoronix Test Suite v10.8.4