OpenCL August Fresh NVIDIA vs. Radeon OpenCL Linux benchmarks. Tests by Michael Larabel for a future article on Phoronix.com. Radeon RX Vega 56: Processor: AMD Ryzen Threadripper 2990WX 32-Core @ 3.00GHz (32 Cores / 64 Threads), Motherboard: ASUS ROG ZENITH EXTREME (1402 BIOS), Chipset: AMD Family 17h, Memory: 32768MB, Disk: Samsung SSD 970 EVO 500GB, Graphics: AMD Radeon RX Vega 8176MB, Audio: Realtek ALC1220, Monitor: ASUS VP28U, Network: Intel I211 Gigabit Connection + Qualcomm Atheros QCA6174 802.11ac Wireless OS: Ubuntu 18.04, Kernel: 4.15.0-33-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: amdgpu 18.0.99, OpenGL: 4.6.13536, OpenCL: OpenCL 2.1 AMD-APP (2671.3), Compiler: GCC 7.3.0, File-System: ext4, Screen Resolution: 3840x2160 Radeon RX Vega 64: Processor: AMD Ryzen Threadripper 2990WX 32-Core @ 3.00GHz (32 Cores / 64 Threads), Motherboard: ASUS ROG ZENITH EXTREME (1402 BIOS), Chipset: AMD Family 17h, Memory: 32768MB, Disk: Samsung SSD 970 EVO 500GB, Graphics: AMD Radeon RX Vega 8176MB, Audio: Realtek ALC1220, Monitor: ASUS VP28U, Network: Intel I211 Gigabit Connection + Qualcomm Atheros QCA6174 802.11ac Wireless OS: Ubuntu 18.04, Kernel: 4.15.0-33-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: amdgpu 18.0.99, OpenGL: 4.6.13536, OpenCL: OpenCL 2.1 AMD-APP (2671.3), Compiler: GCC 7.3.0, File-System: ext4, Screen Resolution: 3840x2160 GeForce GTX 1070: Processor: AMD Ryzen Threadripper 2990WX 32-Core @ 3.00GHz (32 Cores / 64 Threads), Motherboard: ASUS ROG ZENITH EXTREME (1402 BIOS), Chipset: AMD Family 17h, Memory: 32768MB, Disk: Samsung SSD 970 EVO 500GB, Graphics: NVIDIA GeForce GTX 1070 8192MB (1506/4006MHz), Audio: Realtek ALC1220, Monitor: ASUS VP28U, Network: Intel I211 Gigabit Connection + Qualcomm Atheros QCA6174 802.11ac Wireless OS: Ubuntu 18.04, Kernel: 4.15.0-33-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: NVIDIA 396.54, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 9.2.210, Compiler: GCC 7.3.0, File-System: ext4, Screen Resolution: 3840x2160 GeForce GTX 1070 Ti: Processor: AMD Ryzen Threadripper 2990WX 32-Core @ 3.00GHz (32 Cores / 64 Threads), Motherboard: ASUS ROG ZENITH EXTREME (1402 BIOS), Chipset: AMD Family 17h, Memory: 32768MB, Disk: Samsung SSD 970 EVO 500GB, Graphics: Zotac NVIDIA GeForce GTX 1070 Ti 8192MB (1607/4006MHz), Audio: Realtek ALC1220, Monitor: ASUS VP28U, Network: Intel I211 Gigabit Connection + Qualcomm Atheros QCA6174 802.11ac Wireless OS: Ubuntu 18.04, Kernel: 4.15.0-33-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: NVIDIA 396.54, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 9.2.210, Compiler: GCC 7.3.0, File-System: ext4, Screen Resolution: 3840x2160 GeForce GTX 1080: Processor: AMD Ryzen Threadripper 2990WX 32-Core @ 3.00GHz (32 Cores / 64 Threads), Motherboard: ASUS ROG ZENITH EXTREME (1402 BIOS), Chipset: AMD Family 17h, Memory: 32768MB, Disk: Samsung SSD 970 EVO 500GB, Graphics: NVIDIA GeForce GTX 1080 8192MB (1607/5005MHz), Audio: Realtek ALC1220, Monitor: ASUS VP28U, Network: Intel I211 Gigabit Connection + Qualcomm Atheros QCA6174 802.11ac Wireless OS: Ubuntu 18.04, Kernel: 4.15.0-33-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: NVIDIA 396.54, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 9.2.210, Compiler: GCC 7.3.0, File-System: ext4, Screen Resolution: 3840x2160 GeForce GTX 1080 Ti: Processor: AMD Ryzen Threadripper 2990WX 32-Core @ 3.00GHz (32 Cores / 64 Threads), Motherboard: ASUS ROG ZENITH EXTREME (1402 BIOS), Chipset: AMD Family 17h, Memory: 32768MB, Disk: Samsung SSD 970 EVO 500GB, Graphics: NVIDIA GeForce GTX 1080 Ti 11264MB (1480/5508MHz), Audio: Realtek ALC1220, Monitor: ASUS VP28U, Network: Intel I211 Gigabit Connection + Qualcomm Atheros QCA6174 802.11ac Wireless OS: Ubuntu 18.04, Kernel: 4.15.0-33-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: NVIDIA 396.54, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 9.2.210, Compiler: GCC 7.3.0, File-System: ext4, Screen Resolution: 3840x2160 Radeon R7 250X + Radeon R7 260X: Processor: AMD FX-6300 Six-Core @ 3.50GHz (3 Cores / 6 Threads), Motherboard: MSI 970 GAMING (MS-7693) v4.0 (V22.4 BIOS), Chipset: AMD RD9x0/RX980, Memory: 16GB, Disk: 1000GB Western Digital WD10EZEX-08W + 275GB Crucial CT275MX3 + 2000GB Seagate ST2000DX002-2DV1 + 64GB Cruzer Blade, Graphics: Sapphire AMD Radeon HD 7770/8760 / R7 250X 2GB, Audio: Realtek ALC1150, Monitor: BenQ GW2270, Network: Qualcomm Atheros AR9485 OS: openSUSE 20210708, Kernel: 5.13.0-1-default (x86_64), Desktop: KDE Plasma 5.22.2, Display Server: X Server 1.20.11, Display Driver: modesetting 1.20.11, OpenGL: 4.6 Mesa 21.1.4 (LLVM 12.0.0), OpenCL: OpenCL 2.1 AMD-APP (3224.4), Vulkan: 1.2.168, Compiler: GCC 11.1.1 20210625 [revision 62bbb113ae68a7e724255e17143520735bcb9ec9] + Clang 12.0.0Target: + LLVM 12.0.0, File-System: ext4, Screen Resolution: 1920x1080 SHOC Scalable HeterOgeneous Computing 2015-11-10 Target: OpenCL - Benchmark: FFT SP GFLOPS > Higher Is Better Radeon RX Vega 56 ............... 830.77 |=============================== Radeon RX Vega 64 ............... 882.96 |================================= GeForce GTX 1070 ................ 528.71 |==================== GeForce GTX 1070 Ti ............. 551.95 |===================== GeForce GTX 1080 ................ 650.78 |======================== GeForce GTX 1080 Ti ............. 984.61 |===================================== Radeon R7 250X + Radeon R7 260X . 149.01 |====== SHOC Scalable HeterOgeneous Computing 2015-11-10 Target: OpenCL - Benchmark: MD5 Hash GHash/s > Higher Is Better Radeon RX Vega 56 ............... 13.1000 |======================== Radeon RX Vega 64 ............... 17.1200 |=============================== GeForce GTX 1070 ................ 10.7000 |==================== GeForce GTX 1070 Ti ............. 13.8000 |========================= GeForce GTX 1080 ................ 14.4000 |========================== GeForce GTX 1080 Ti ............. 19.7200 |==================================== Radeon R7 250X + Radeon R7 260X . 1.5159 |=== SHOC Scalable HeterOgeneous Computing 2015-11-10 Target: OpenCL - Benchmark: Texture Read Bandwidth GB/s > Higher Is Better Radeon RX Vega 56 ............... 362.97 |======================= Radeon RX Vega 64 ............... 427.74 |=========================== GeForce GTX 1070 ................ 456.10 |============================ GeForce GTX 1070 Ti ............. 501.46 |=============================== GeForce GTX 1080 ................ 523.65 |================================= GeForce GTX 1080 Ti ............. 593.14 |===================================== Radeon R7 250X + Radeon R7 260X . 68.59 |==== cl-mem 2017-01-13 Benchmark: Copy GB/s > Higher Is Better Radeon RX Vega 56 ............... 313.30 |=============================== Radeon RX Vega 64 ............... 369.93 |===================================== GeForce GTX 1070 ................ 186.87 |=================== GeForce GTX 1070 Ti ............. 186.80 |=================== GeForce GTX 1080 ................ 209.33 |===================== GeForce GTX 1080 Ti ............. 317.37 |================================ Radeon R7 250X + Radeon R7 260X . 45.90 |===== cl-mem 2017-01-13 Benchmark: Read GB/s > Higher Is Better Radeon RX Vega 56 ............... 346.47 |================================ Radeon RX Vega 64 ............... 399.00 |===================================== GeForce GTX 1070 ................ 205.50 |=================== GeForce GTX 1070 Ti ............. 205.63 |=================== GeForce GTX 1080 ................ 228.40 |===================== GeForce GTX 1080 Ti ............. 337.73 |=============================== Radeon R7 250X + Radeon R7 260X . 63.40 |====== cl-mem 2017-01-13 Benchmark: Write GB/s > Higher Is Better Radeon RX Vega 56 ............... 333.20 |================================ Radeon RX Vega 64 ............... 388.57 |===================================== GeForce GTX 1070 ................ 192.30 |================== GeForce GTX 1070 Ti ............. 191.00 |================== GeForce GTX 1080 ................ 216.70 |===================== GeForce GTX 1080 Ti ............. 336.30 |================================ Radeon R7 250X + Radeon R7 260X . 39.20 |==== FAHBench 2.3.2 Ns Per Day > Higher Is Better Radeon RX Vega 56 ............... 92.48 |=================== Radeon RX Vega 64 ............... 92.46 |=================== GeForce GTX 1070 ................ 131.18 |=========================== GeForce GTX 1070 Ti ............. 145.38 |============================== GeForce GTX 1080 ................ 141.38 |============================= GeForce GTX 1080 Ti ............. 179.34 |===================================== Radeon R7 250X + Radeon R7 260X . 19.05 |==== MandelGPU 1.3pts1 OpenCL Device: GPU Samples/sec > Higher Is Better Radeon RX Vega 56 ............... 165896206.17 |===================== Radeon RX Vega 64 ............... 192917451.23 |======================== GeForce GTX 1070 ................ 148507150.87 |================== GeForce GTX 1070 Ti ............. 184334944.10 |======================= GeForce GTX 1080 ................ 188560526.97 |======================= GeForce GTX 1080 Ti ............. 250678650.87 |=============================== Radeon R7 250X + Radeon R7 260X . 29262082.90 |==== LuxMark 3.1 OpenCL Device: GPU - Scene: Hotel Score > Higher Is Better Radeon RX Vega 56 ............... 5083 |================================== Radeon RX Vega 64 ............... 5907 |======================================= GeForce GTX 1070 ................ 3820 |========================= GeForce GTX 1070 Ti ............. 4405 |============================= GeForce GTX 1080 ................ 3883 |========================== GeForce GTX 1080 Ti ............. 5662 |===================================== Radeon R7 250X + Radeon R7 260X . 1628 |===========