pro-vs-rocm AMD Ryzen 7 2700X Eight-Core testing with a ASRock X470 Taichi Ultimate (P1.60 BIOS) and AMD Radeon RX Vega 8GB on Fedora 32 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2006169-AS-PROVSROCM52 .
pro-vs-rocm Processor Motherboard Chipset Memory Disk Graphics Audio Monitor Network OS Kernel Desktop Display Server Display Driver OpenGL OpenCL Compiler File-System Screen Resolution pro rocm amdgpu-pro full stack AMD Ryzen 7 2700X Eight-Core @ 3.70GHz (8 Cores / 16 Threads) ASRock X470 Taichi Ultimate (P1.60 BIOS) AMD 17h 32768MB 2 x Samsung SSD 970 PRO 512GB + 500GB Samsung SSD 860 + 2 x 5001GB TOSHIBA HDWE150 + 1000GB Samsung SSD 860 + 500GB Samsung SSD 850 AMD Radeon RX 56/64 8GB (1630/945MHz) AMD Vega 10 HDMI Audio 31MU97 Intel I211 + Aquantia AQC107 NBase-T/IEEE + Intel Dual Band-AC 3168NGW Fedora 32 5.6.16-300.fc32.x86_64 (x86_64) GNOME Shell 3.36.3 X Server 1.20.8 + Wayland modesetting 1.20.8 4.6 Mesa 20.0.7 (LLVM 10.0.0) OpenCL 2.1 AMD-APP (3110.6) GCC 10.1.1 20200507 + Clang 10.0.0 ext4 4096x2160 OpenCL 2.1 AMD-APP (3098.0) AMD Radeon RX Vega 8GB (1630/945MHz) 5.6.18-300.fc32.x86_64 (x86_64) X Server 1.20.8 4.6.13572 OpenCL 2.1 AMD-APP (2906.7) OpenBenchmarking.org Processor Details - Scaling Governor: acpi-cpufreq ondemand Security Details - pro: SELinux + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional STIBP: disabled RSB filling + tsx_async_abort: Not affected - rocm: SELinux + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional STIBP: disabled RSB filling + tsx_async_abort: Not affected - amdgpu-pro full stack: SELinux + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional STIBP: disabled RSB filling + srbds: Not affected + tsx_async_abort: Not affected
pro-vs-rocm darktable: Boat - OpenCL darktable: Boat - CPU-only darktable: Masskrug - OpenCL darktable: Masskrug - CPU-only darktable: Server Rack - OpenCL darktable: Server Room - OpenCL darktable: Server Rack - CPU-only darktable: Server Room - CPU-only pro rocm amdgpu-pro full stack 2.96 13.53 5.04 6.92 0.12 0.87 0.18 4.41 3.96 5.97 0.35 2.45 2.90 5.21 0.16 0.90 OpenBenchmarking.org
Darktable Test: Boat - Acceleration: OpenCL OpenBenchmarking.org Seconds, Fewer Is Better Darktable 3.0.2 Test: Boat - Acceleration: OpenCL pro rocm amdgpu-pro full stack 0.891 1.782 2.673 3.564 4.455 SE +/- 0.03, N = 15 SE +/- 0.09, N = 15 SE +/- 0.04, N = 12 2.96 3.96 2.90
Darktable Test: Boat - Acceleration: CPU-only OpenBenchmarking.org Seconds, Fewer Is Better Darktable 3.0.2 Test: Boat - Acceleration: CPU-only pro 3 6 9 12 15 SE +/- 0.02, N = 3 13.53
Darktable Test: Masskrug - Acceleration: OpenCL OpenBenchmarking.org Seconds, Fewer Is Better Darktable 3.0.2 Test: Masskrug - Acceleration: OpenCL pro rocm amdgpu-pro full stack 1.3433 2.6866 4.0299 5.3732 6.7165 SE +/- 0.06, N = 3 SE +/- 0.07, N = 15 SE +/- 0.04, N = 3 5.04 5.97 5.21
Darktable Test: Masskrug - Acceleration: CPU-only OpenBenchmarking.org Seconds, Fewer Is Better Darktable 3.0.2 Test: Masskrug - Acceleration: CPU-only pro 2 4 6 8 10 SE +/- 0.03, N = 3 6.92
Darktable Test: Server Rack - Acceleration: OpenCL OpenBenchmarking.org Seconds, Fewer Is Better Darktable 3.0.2 Test: Server Rack - Acceleration: OpenCL pro rocm amdgpu-pro full stack 0.0788 0.1576 0.2364 0.3152 0.394 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 0.12 0.35 0.16
Darktable Test: Server Room - Acceleration: OpenCL OpenBenchmarking.org Seconds, Fewer Is Better Darktable 3.0.2 Test: Server Room - Acceleration: OpenCL pro rocm amdgpu-pro full stack 0.5513 1.1026 1.6539 2.2052 2.7565 SE +/- 0.01, N = 5 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 0.87 2.45 0.90
Darktable Test: Server Rack - Acceleration: CPU-only OpenBenchmarking.org Seconds, Fewer Is Better Darktable 3.0.2 Test: Server Rack - Acceleration: CPU-only pro 0.0405 0.081 0.1215 0.162 0.2025 SE +/- 0.00, N = 10 0.18
Darktable Test: Server Room - Acceleration: CPU-only OpenBenchmarking.org Seconds, Fewer Is Better Darktable 3.0.2 Test: Server Room - Acceleration: CPU-only pro 0.9923 1.9846 2.9769 3.9692 4.9615 SE +/- 0.02, N = 3 4.41
Phoronix Test Suite v10.8.4