OpenCL Radeon Linux AMD Benchmarks by Michael Larabel. AMD Ryzen 9 7950X 16-Core testing with a ASUS ROG CROSSHAIR X670E HERO (9922 BIOS) and Sapphire AMD Radeon RX 6500 XT 4GB on Ubuntu 23.04 via the Phoronix Test Suite. RX 6600: Processor: AMD Ryzen 9 7950X 16-Core @ 4.50GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR X670E HERO (9922 BIOS), Chipset: AMD Device 14d8, Memory: 2 x 16 GB DDR5-6000MT/s F5-6000J3038F16G, Disk: Western Digital WD_BLACK SN850X 1000GB + 2000GB, Graphics: Gigabyte AMD Radeon RX 6600 8GB (2750/875MHz), Audio: AMD Navi 21/23, Monitor: ASUS MG28U, Network: Intel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411 OS: Ubuntu 23.04, Kernel: 6.2.2-060202-generic (x86_64), Desktop: GNOME Shell 43.2, Display Server: X Server 1.21.1.6, OpenGL: 4.6 Mesa 23.1.0-devel (git-5f5e30b 2023-03-09 lunar-oibaf-ppa) (LLVM 15.0.7 DRM 3.49), OpenCL: OpenCL 2.1 AMD-APP (3513.0), Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 3840x2160 RX 6700 XT: Processor: AMD Ryzen 9 7950X 16-Core @ 4.50GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR X670E HERO (9922 BIOS), Chipset: AMD Device 14d8, Memory: 2 x 16 GB DDR5-6000MT/s F5-6000J3038F16G, Disk: Western Digital WD_BLACK SN850X 1000GB + 2000GB, Graphics: AMD Radeon RX 6700 XT 12GB (2855/1000MHz), Audio: AMD Navi 21/23, Monitor: ASUS MG28U, Network: Intel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411 OS: Ubuntu 23.04, Kernel: 6.2.2-060202-generic (x86_64), Desktop: GNOME Shell 43.2, Display Server: X Server 1.21.1.6, OpenGL: 4.6 Mesa 23.1.0-devel (git-5f5e30b 2023-03-09 lunar-oibaf-ppa) (LLVM 15.0.7 DRM 3.49), OpenCL: OpenCL 2.1 AMD-APP (3513.0), Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 3840x2160 RX 6500 XT: Processor: AMD Ryzen 9 7950X 16-Core @ 4.50GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR X670E HERO (9922 BIOS), Chipset: AMD Device 14d8, Memory: 2 x 16 GB DDR5-6000MT/s F5-6000J3038F16G, Disk: Western Digital WD_BLACK SN850X 1000GB + 2000GB, Graphics: Sapphire AMD Radeon RX 6500 XT 4GB (2975/1124MHz), Audio: AMD Navi 21/23, Monitor: ASUS MG28U, Network: Intel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411 OS: Ubuntu 23.04, Kernel: 6.2.2-060202-generic (x86_64), Desktop: GNOME Shell 43.2, Display Server: X Server 1.21.1.6, OpenGL: 4.6 Mesa 23.1.0-devel (git-5f5e30b 2023-03-09 lunar-oibaf-ppa) (LLVM 15.0.7 DRM 3.49), OpenCL: OpenCL 2.1 AMD-APP (3513.0), Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 3840x2160 GPU Power Consumption Monitor Phoronix Test Suite System Monitoring Watts RX 6600 .... MIN: 3 AVG: 39 MAX: 100 RX 6700 XT . MIN: 4 AVG: 67 MAX: 188 SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Triad GB/s > Higher Is Better RX 6600 .... 12.2654 |============================== RX 6700 XT . 23.6167 |========================================================= RX 6500 XT . 6.3591 |=============== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Triad GB/s Per Watt > Higher Is Better RX 6600 .... 2.125 |==================================== RX 6700 XT . 3.497 |=========================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Reduction GB/s > Higher Is Better RX 6600 .... 208.08 |=================== RX 6700 XT . 626.31 |========================================================== RX 6500 XT . 139.27 |============= SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Reduction GB/s Per Watt > Higher Is Better RX 6600 .... 20.34 |================= RX 6700 XT . 70.82 |=========================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Bus Speed Download GB/s > Higher Is Better RX 6600 .... 14.3363 |============================ RX 6700 XT . 28.8561 |========================================================= RX 6500 XT . 7.1636 |============== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Bus Speed Download GB/s Per Watt > Higher Is Better RX 6600 .... 2.789 |===================================== RX 6700 XT . 4.461 |=========================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Bus Speed Readback GB/s > Higher Is Better RX 6600 .... 14.0904 |============================== RX 6700 XT . 26.4047 |========================================================= RX 6500 XT . 7.0477 |=============== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Bus Speed Readback GB/s Per Watt > Higher Is Better RX 6600 .... 2.662 |==================================== RX 6700 XT . 4.355 |=========================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Texture Read Bandwidth GB/s > Higher Is Better RX 6600 .... 603.59 |======================================================= RX 6700 XT . 641.16 |========================================================== RX 6500 XT . 535.11 |================================================ SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Texture Read Bandwidth GB/s Per Watt > Higher Is Better RX 6600 .... 38.22 |=========================================================== RX 6700 XT . 33.83 |==================================================== clpeak 1.1.2 OpenCL Test: Global Memory Bandwidth GBPS Per Watt > Higher Is Better RX 6600 .... 6.714 |============================================ RX 6700 XT . 9.043 |=========================================================== clpeak 1.1.2 OpenCL Test: Transfer Bandwidth enqueueReadBuffer GBPS Per Watt > Higher Is Better RX 6600 .... 1.560 |=========================================================== RX 6700 XT . 1.229 |============================================== clpeak 1.1.2 OpenCL Test: Global Memory Bandwidth GBPS > Higher Is Better RX 6600 .... 191.26 |==================================== RX 6700 XT . 311.69 |========================================================== RX 6500 XT . 122.86 |======================= clpeak 1.1.2 OpenCL Test: Transfer Bandwidth enqueueReadBuffer GBPS > Higher Is Better RX 6600 .... 4.97 |============================================================ RX 6700 XT . 4.99 |============================================================ RX 6500 XT . 5.01 |============================================================ clpeak 1.1.2 OpenCL Test: Transfer Bandwidth enqueueWriteBuffer GBPS > Higher Is Better RX 6600 .... 22.64 |========================================================= RX 6700 XT . 22.65 |========================================================= RX 6500 XT . 23.52 |=========================================================== clpeak 1.1.2 OpenCL Test: Transfer Bandwidth enqueueWriteBuffer GBPS Per Watt > Higher Is Better RX 6600 .... 7.173 |=========================================================== RX 6700 XT . 5.644 |============================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: GEMM SGEMM_N GFLOPS Per Watt > Higher Is Better RX 6600 .... 102.61 |========================== RX 6700 XT . 228.49 |========================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: S3D GFLOPS > Higher Is Better RX 6600 .... 79.76 |========================================== RX 6700 XT . 110.30 |========================================================== RX 6500 XT . 23.54 |============ clpeak 1.1.2 OpenCL Test: Double-Precision Compute GFLOPS > Higher Is Better RX 6600 .... 569.74 |========================================= RX 6700 XT . 807.53 |========================================================== RX 6500 XT . 336.00 |======================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: S3D GFLOPS Per Watt > Higher Is Better RX 6600 .... 12.16 |========================================================= RX 6700 XT . 12.52 |=========================================================== clpeak 1.1.2 OpenCL Test: Double-Precision Compute GFLOPS Per Watt > Higher Is Better RX 6600 .... 34.86 |========================================================= RX 6700 XT . 35.80 |=========================================================== clpeak 1.1.2 OpenCL Test: Single-Precision Compute GFLOPS Per Watt > Higher Is Better RX 6600 .... 839.62 |===================================================== RX 6700 XT . 916.94 |========================================================== clpeak 1.1.2 OpenCL Test: Single-Precision Compute GFLOPS > Higher Is Better RX 6600 .... 8032.82 |======================================= RX 6700 XT . 11441.23 |======================================================== RX 6500 XT . 4790.97 |======================= SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: GEMM SGEMM_N GFLOPS > Higher Is Better RX 6600 .... 1778.07 |======================= RX 6700 XT . 4337.60 |========================================================= RX 6500 XT . 1063.03 |============== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: MD5 Hash GHash/s > Higher Is Better RX 6600 .... 11.8971 |========================================= RX 6700 XT . 16.3937 |========================================================= RX 6500 XT . 7.0820 |========================= SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: MD5 Hash GHash/s Per Watt > Higher Is Better RX 6600 .... 1.147 |=========================================================== RX 6700 XT . 1.002 |==================================================== clpeak 1.1.2 OpenCL Test: Integer Compute GIOPS > Higher Is Better RX 6600 .... 2164.23 |========================================= RX 6700 XT . 2991.01 |========================================================= RX 6500 XT . 1270.59 |======================== clpeak 1.1.2 OpenCL Test: Integer Compute GIOPS Per Watt > Higher Is Better RX 6600 .... 142.02 |========================================================== RX 6700 XT . 139.75 |========================================================= clpeak 1.1.2 OpenCL Test: Integer 24-bit Compute GIOPS > Higher Is Better RX 6600 .... 7831.73 |========================================= RX 6700 XT . 10802.43 |======================================================== RX 6500 XT . 4680.82 |======================== clpeak 1.1.2 OpenCL Test: Integer 24-bit Compute GIOPS Per Watt > Higher Is Better RX 6600 .... 938.10 |========================================================= RX 6700 XT . 954.02 |========================================================== FluidX3D 2.3 Test: FP32-FP32 MLUPs/s Per Watt > Higher Is Better RX 6600 .... 16.11 |=========================================================== RX 6700 XT . 13.82 |=================================================== FluidX3D 2.3 Test: FP32-FP32 MLUPs/s > Higher Is Better RX 6600 .... 962 |========================================== RX 6700 XT . 1382 |============================================================ RX 6500 XT . 495 |===================== FluidX3D 2.3 Test: FP32-FP16C MLUPs/s > Higher Is Better RX 6600 .... 1838 |======================================= RX 6700 XT . 2793 |============================================================ RX 6500 XT . 1030 |====================== FluidX3D 2.3 Test: FP32-FP16S MLUPs/s > Higher Is Better RX 6600 .... 1816 |======================================= RX 6700 XT . 2786 |============================================================ RX 6500 XT . 1011 |====================== FluidX3D 2.3 Test: FP32-FP16C MLUPs/s Per Watt > Higher Is Better RX 6600 .... 22.52 |=========================================================== RX 6700 XT . 20.86 |======================================================= FluidX3D 2.3 Test: FP32-FP16S MLUPs/s Per Watt > Higher Is Better RX 6600 .... 28.09 |=========================================================== RX 6700 XT . 26.03 |======================================================= LeelaChessZero 0.28 Backend: OpenCL Nodes Per Second > Higher Is Better RX 6600 .... 14791 |============================================= RX 6700 XT . 19460 |=========================================================== RX 6500 XT . 7390 |====================== LeelaChessZero 0.28 Backend: OpenCL Nodes Per Second Per Watt > Higher Is Better RX 6600 .... 152.26 |========================================================== RX 6700 XT . 115.31 |============================================ clpeak 1.1.2 OpenCL Test: Kernel Latency us < Lower Is Better RX 6600 .... 11.80 |==================================================== RX 6700 XT . 13.39 |=========================================================== RX 6500 XT . 11.45 |================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3.0 AVG: 6.6 MAX: 15.0 RX 6700 XT . MIN: 4.0 AVG: 8.8 MAX: 21.0 FluidX3D 2.3 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3 AVG: 60 MAX: 65 |--------------------X| RX 6700 XT . MIN: 4 AVG: 100 MAX: 110 FluidX3D 2.3 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3 AVG: 82 MAX: 91 .|--------------------X RX 6700 XT . MIN: 5 AVG: 134 MAX: 156 FluidX3D 2.3 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3 AVG: 65 MAX: 72 |--------------------X RX 6700 XT . MIN: 4 AVG: 107 MAX: 124 clpeak 1.1.2 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3.0 AVG: 4.8 MAX: 19.0 RX 6700 XT . MIN: 4.0 AVG: 6.1 MAX: 26.0 clpeak 1.1.2 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3 AVG: 15 MAX: 100 RX 6700 XT . MIN: 4 AVG: 21 MAX: 142 clpeak 1.1.2 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3 AVG: 8 MAX: 100 RX 6700 XT . MIN: 4 AVG: 11 MAX: 161 clpeak 1.1.2 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3 AVG: 28 MAX: 72 X####################-| RX 6700 XT . MIN: 4 AVG: 34 MAX: 123 clpeak 1.1.2 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3.0 AVG: 16.3 MAX: 80.0 RX 6700 XT . MIN: 4.0 AVG: 22.6 MAX: 97.0 clpeak 1.1.2 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3 AVG: 10 MAX: 94 X-----------------------| . RX 6700 XT . MIN: 4 AVG: 12 MAX: 135 clpeak 1.1.2 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3.0 AVG: 3.2 MAX: 8.0 RX 6700 XT . MIN: 4.0 AVG: 4.1 MAX: 11.0 clpeak 1.1.2 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3.0 AVG: 3.2 MAX: 8.0 RX 6700 XT . MIN: 4.0 AVG: 4.0 MAX: 5.0 LeelaChessZero 0.28 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3 AVG: 97 MAX: 100 RX 6700 XT . MIN: 4 AVG: 169 MAX: 188 SHOC Scalable HeterOgeneous Computing 2020-04-17 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3.0 AVG: 5.8 MAX: 24.0 RX 6700 XT . MIN: 5.0 AVG: 6.8 MAX: 28.0 SHOC Scalable HeterOgeneous Computing 2020-04-17 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3 AVG: 10 MAX: 100 RX 6700 XT . MIN: 4 AVG: 16 MAX: 176 SHOC Scalable HeterOgeneous Computing 2020-04-17 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 4.0 AVG: 10.2 MAX: 65.0 RX 6700 XT . MIN: 4.0 AVG: 8.8 MAX: 79.0 SHOC Scalable HeterOgeneous Computing 2020-04-17 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3 AVG: 17 MAX: 99 X--------------------| RX 6700 XT . MIN: 4 AVG: 19 MAX: 174 SHOC Scalable HeterOgeneous Computing 2020-04-17 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3.0 AVG: 5.1 MAX: 31.0 RX 6700 XT . MIN: 4.0 AVG: 6.5 MAX: 44.0 SHOC Scalable HeterOgeneous Computing 2020-04-17 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3.0 AVG: 5.3 MAX: 32.0 RX 6700 XT . MIN: 4.0 AVG: 6.1 MAX: 42.0 SHOC Scalable HeterOgeneous Computing 2020-04-17 GPU Power Consumption Monitor Watts < Lower Is Better RX 6600 .... MIN: 3.0 AVG: 15.8 MAX: 83.0 RX 6700 XT . MIN: 4.0 AVG: 19.0 MAX: 84.0