ngc rtx 3090 AMD Ryzen 9 5900X 12-Core testing with a ASUS ROG CROSSHAIR VIII HERO (3402 BIOS) and NVIDIA GeForce RTX 3090 24GB on Ubuntu 20.04 via the Phoronix Test Suite. 1: Processor: AMD Ryzen 9 5900X 12-Core @ 3.70GHz (12 Cores / 24 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (3402 BIOS), Chipset: AMD Starship/Matisse, Memory: 16GB, Disk: 1000GB Sabrent Rocket 4.0 Plus + 2000GB, Graphics: NVIDIA GeForce RTX 3090 24GB, Audio: NVIDIA Device 1aef, Monitor: ASUS VP28U, Network: Realtek RTL8125 2.5GbE + Intel I211 OS: Ubuntu 20.04, Kernel: 5.8.0-48-generic (x86_64), Desktop: GNOME Shell 3.36.7, Display Server: X Server 1.20.9, Display Driver: NVIDIA 460.67, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 11.2.162, Vulkan: 1.2.155, Compiler: GCC 9.3.0 + CUDA 11.2, File-System: ext4, Screen Resolution: 3840x2160 2: Processor: AMD Ryzen 9 5900X 12-Core @ 3.70GHz (12 Cores / 24 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (3402 BIOS), Chipset: AMD Starship/Matisse, Memory: 16GB, Disk: 1000GB Sabrent Rocket 4.0 Plus + 2000GB, Graphics: NVIDIA GeForce RTX 3090 24GB, Audio: NVIDIA Device 1aef, Monitor: ASUS VP28U, Network: Realtek RTL8125 2.5GbE + Intel I211 OS: Ubuntu 20.04, Kernel: 5.8.0-48-generic (x86_64), Desktop: GNOME Shell 3.36.7, Display Server: X Server 1.20.9, Display Driver: NVIDIA 460.67, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 11.2.162, Vulkan: 1.2.155, Compiler: GCC 9.3.0 + CUDA 11.2, File-System: ext4, Screen Resolution: 3840x2160 3: Processor: AMD Ryzen 9 5900X 12-Core @ 3.70GHz (12 Cores / 24 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (3402 BIOS), Chipset: AMD Starship/Matisse, Memory: 16GB, Disk: 1000GB Sabrent Rocket 4.0 Plus + 2000GB, Graphics: NVIDIA GeForce RTX 3090 24GB, Audio: NVIDIA Device 1aef, Monitor: ASUS VP28U, Network: Realtek RTL8125 2.5GbE + Intel I211 OS: Ubuntu 20.04, Kernel: 5.8.0-48-generic (x86_64), Desktop: GNOME Shell 3.36.7, Display Server: X Server 1.20.9, Display Driver: NVIDIA 460.67, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 11.2.162, Vulkan: 1.2.155, Compiler: GCC 9.3.0 + CUDA 11.2, File-System: ext4, Screen Resolution: 3840x2160 RealSR-NCNN 20200818 Scale: 4x - TAA: No Seconds < Lower Is Better 1 . 5.866 |==================================================================== 2 . 5.867 |==================================================================== 3 . 5.845 |==================================================================== RealSR-NCNN 20200818 Scale: 4x - TAA: Yes Seconds < Lower Is Better 1 . 29.62 |==================================================================== 2 . 29.60 |==================================================================== 3 . 29.57 |==================================================================== Waifu2x-NCNN Vulkan 20200818 Scale: 2x - Denoise: 3 - TAA: Yes Seconds < Lower Is Better 1 . 3.439 |==================================================================== 2 . 3.444 |==================================================================== 3 . 3.440 |==================================================================== VkFFT 1.1.1 Benchmark Score > Higher Is Better 1 . 43686 |================================================================== 2 . 43194 |================================================================== 3 . 44750 |==================================================================== Hashcat 6.1.1 Benchmark: MD5 H/s > Higher Is Better 1 . 66413233333 |============================================================== 2 . 66432866667 |============================================================== 3 . 66617033333 |============================================================== Hashcat 6.1.1 Benchmark: SHA1 H/s > Higher Is Better 1 . 22583233333 |============================================================== 2 . 22585700000 |============================================================== 3 . 22665866667 |============================================================== Hashcat 6.1.1 Benchmark: 7-Zip H/s > Higher Is Better 1 . 1149333 |================================================================== 2 . 1144667 |================================================================== 3 . 1147133 |================================================================== Hashcat 6.1.1 Benchmark: SHA-512 H/s > Higher Is Better 1 . 2853033333 |=============================================================== 2 . 2851266667 |=============================================================== 3 . 2857366667 |=============================================================== Hashcat 6.1.1 Benchmark: TrueCrypt RIPEMD160 + XTS H/s > Higher Is Better 1 . 846967 |=================================================================== 2 . 845867 |=================================================================== 3 . 849167 |=================================================================== Mixbench 2020-06-23 Backend: OpenCL - Benchmark: Integer GIOPS > Higher Is Better 1 . 20685.77 |================================================================= 2 . 19683.19 |============================================================== 3 . 20677.83 |================================================================= Mixbench 2020-06-23 Backend: OpenCL - Benchmark: Double Precision GFLOPS > Higher Is Better 1 . 496.29 |=================================================================== 2 . 467.53 |=============================================================== 3 . 459.26 |============================================================== Mixbench 2020-06-23 Backend: OpenCL - Benchmark: Single Precision GFLOPS > Higher Is Better 1 . 34053.47 |============================================================== 2 . 34079.35 |============================================================== 3 . 35867.33 |================================================================= SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: S3D GFLOPS > Higher Is Better 1 . 428.92 |=================================================================== 2 . 430.51 |=================================================================== 3 . 429.95 |=================================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Triad GB/s > Higher Is Better 1 . 25.48 |==================================================================== 2 . 25.46 |==================================================================== 3 . 25.47 |==================================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: FFT SP GFLOPS > Higher Is Better 1 . 2347.21 |================================================================== 2 . 2342.08 |================================================================== 3 . 2344.26 |================================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: MD5 Hash GHash/s > Higher Is Better 1 . 44.19 |==================================================================== 2 . 43.96 |==================================================================== 3 . 44.02 |==================================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Reduction GB/s > Higher Is Better 1 . 390.96 |=================================================================== 2 . 389.90 |=================================================================== 3 . 391.28 |=================================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: GEMM SGEMM_N GFLOPS > Higher Is Better 1 . 8202.75 |================================================================= 2 . 8272.77 |================================================================== 3 . 8229.33 |================================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Max SP Flops GFLOPS > Higher Is Better 1 . 39924.9 |================================================================== 2 . 39799.6 |================================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Bus Speed Download GB/s > Higher Is Better 1 . 26.31 |==================================================================== 2 . 26.31 |==================================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Bus Speed Readback GB/s > Higher Is Better 1 . 26.36 |==================================================================== 2 . 26.39 |==================================================================== SHOC Scalable HeterOgeneous Computing 2020-04-17 Target: OpenCL - Benchmark: Texture Read Bandwidth GB/s > Higher Is Better 1 . 2224.01 |================================================================== 2 . 2226.47 |================================================================== GROMACS 2020.3 Water Benchmark Ns Per Day > Higher Is Better 1 . 9.643 |==================================================================== 2 . 9.670 |==================================================================== cl-mem 2017-01-13 Benchmark: Copy GB/s > Higher Is Better 1 . 363.5 |==================================================================== 2 . 363.0 |==================================================================== cl-mem 2017-01-13 Benchmark: Read GB/s > Higher Is Better 1 . 825.3 |==================================================================== 2 . 825.9 |==================================================================== cl-mem 2017-01-13 Benchmark: Write GB/s > Higher Is Better 1 . 742.0 |==================================================================== 2 . 740.8 |==================================================================== NAMD CUDA 2.14 ATPase Simulation - 327,506 Atoms days/ns < Lower Is Better 1 . 0.12884 |================================================================== 2 . 0.12599 |================================================================= Betsy GPU Compressor 1.1 Beta Codec: ETC1 - Quality: Highest Seconds < Lower Is Better 1 . 3.040 |==================================================================== 2 . 3.015 |=================================================================== Betsy GPU Compressor 1.1 Beta Codec: ETC2 RGB - Quality: Highest Seconds < Lower Is Better 1 . 4.054 |==================================================================== 2 . 4.049 |==================================================================== VkResample 1.0 Upscale: 2x - Precision: Double ms < Lower Is Better 1 . 122.80 |=================================================================== 2 . 122.87 |=================================================================== VkResample 1.0 Upscale: 2x - Precision: Single ms < Lower Is Better 1 . 9.282 |==================================================================== 2 . 9.284 |==================================================================== OctaneBench 2020.1 Total Score Score > Higher Is Better 1 . 680.72 |=================================================================== 2 . 680.19 |=================================================================== RedShift Demo 3.0 Seconds < Lower Is Better 1 . 141 |====================================================================== 2 . 141 |====================================================================== LuxCoreRender OpenCL 2.3 Scene: DLSC M samples/sec > Higher Is Better 1 . 11.20 |==================================================================== 2 . 11.17 |==================================================================== LuxCoreRender OpenCL 2.3 Scene: Food M samples/sec > Higher Is Better 1 . 4.70 |===================================================================== 2 . 4.68 |===================================================================== LuxCoreRender OpenCL 2.3 Scene: LuxCore Benchmark M samples/sec > Higher Is Better 1 . 9.08 |===================================================================== 2 . 9.13 |===================================================================== LuxCoreRender OpenCL 2.3 Scene: Rainbow Colors and Prism M samples/sec > Higher Is Better 1 . 26.12 |==================================================================== 2 . 26.29 |==================================================================== FAHBench 2.3.2 Ns Per Day > Higher Is Better 1 . 343.62 |=================================================================== 2 . 344.72 |=================================================================== LeelaChessZero 0.26 Backend: OpenCL Nodes Per Second > Higher Is Better 1 . 39148 |==================================================================== 2 . 38996 |==================================================================== Rodinia 3.1 Test: OpenCL Particle Filter Seconds < Lower Is Better 1 . 3.743 |==================================================================== 2 . 3.729 |==================================================================== ArrayFire 3.7 Test: Conjugate Gradient OpenCL ms < Lower Is Better 1 . 1.477 |=================================================================== 2 . 1.491 |==================================================================== FinanceBench 2016-07-25 Benchmark: Black-Scholes OpenCL ms < Lower Is Better 1 . 6.234 |==================================================================== 2 . 6.265 |==================================================================== ViennaCL 1.7.1 Test: CPU BLAS - sCOPY GB/s > Higher Is Better 1 . 62.6 |===================================================================== 2 . 62.5 |===================================================================== ViennaCL 1.7.1 Test: CPU BLAS - sAXPY GB/s > Higher Is Better 1 . 91.6 |===================================================================== 2 . 91.5 |===================================================================== ViennaCL 1.7.1 Test: CPU BLAS - sDOT GB/s > Higher Is Better 1 . 136 |===================================================================== 2 . 138 |====================================================================== ViennaCL 1.7.1 Test: CPU BLAS - dCOPY GB/s > Higher Is Better 1 . 22.8 |===================================================================== 2 . 22.5 |==================================================================== ViennaCL 1.7.1 Test: CPU BLAS - dAXPY GB/s > Higher Is Better 1 . 33.4 |===================================================================== 2 . 33.3 |===================================================================== ViennaCL 1.7.1 Test: CPU BLAS - dDOT GB/s > Higher Is Better 1 . 43.8 |===================================================================== 2 . 43.2 |==================================================================== ViennaCL 1.7.1 Test: CPU BLAS - dGEMV-N GB/s > Higher Is Better 1 . 77.8 |===================================================================== 2 . 78.1 |===================================================================== ViennaCL 1.7.1 Test: CPU BLAS - dGEMV-T GB/s > Higher Is Better 1 . 83.2 |===================================================================== 2 . 83.1 |===================================================================== ViennaCL 1.7.1 Test: CPU BLAS - dGEMM-NN GFLOPs/s > Higher Is Better 1 . 53.5 |===================================================================== 2 . 53.2 |===================================================================== ViennaCL 1.7.1 Test: CPU BLAS - dGEMM-NT GFLOPs/s > Higher Is Better 1 . 52.9 |===================================================================== 2 . 52.9 |===================================================================== ViennaCL 1.7.1 Test: CPU BLAS - dGEMM-TN GFLOPs/s > Higher Is Better 1 . 56.7 |===================================================================== 2 . 56.5 |===================================================================== ViennaCL 1.7.1 Test: CPU BLAS - dGEMM-TT GFLOPs/s > Higher Is Better 1 . 55.3 |===================================================================== 2 . 55.1 |===================================================================== ViennaCL 1.7.1 Test: OpenCL BLAS - sCOPY GB/s > Higher Is Better 1 . 366 |====================================================================== 2 . 364 |====================================================================== ViennaCL 1.7.1 Test: OpenCL BLAS - sAXPY GB/s > Higher Is Better 1 . 504 |====================================================================== 2 . 503 |====================================================================== ViennaCL 1.7.1 Test: OpenCL BLAS - sDOT GB/s > Higher Is Better 1 . 375 |====================================================================== 2 . 374 |====================================================================== ViennaCL 1.7.1 Test: OpenCL BLAS - dCOPY GB/s > Higher Is Better 1 . 607 |====================================================================== 2 . 607 |====================================================================== ViennaCL 1.7.1 Test: OpenCL BLAS - dAXPY GB/s > Higher Is Better 1 . 722 |====================================================================== 2 . 722 |====================================================================== ViennaCL 1.7.1 Test: OpenCL BLAS - dDOT GB/s > Higher Is Better 1 . 650 |====================================================================== 2 . 651 |====================================================================== ViennaCL 1.7.1 Test: OpenCL BLAS - dGEMV-N GB/s > Higher Is Better 1 . 237 |====================================================================== 2 . 237 |====================================================================== ViennaCL 1.7.1 Test: OpenCL BLAS - dGEMV-T GB/s > Higher Is Better 1 . 378 |====================================================================== 2 . 376 |====================================================================== ViennaCL 1.7.1 Test: OpenCL BLAS - dGEMM-NN GFLOPs/s > Higher Is Better 1 . 603 |====================================================================== 2 . 602 |====================================================================== ViennaCL 1.7.1 Test: OpenCL BLAS - dGEMM-NT GFLOPs/s > Higher Is Better 1 . 605 |====================================================================== 2 . 604 |====================================================================== ViennaCL 1.7.1 Test: OpenCL BLAS - dGEMM-TN GFLOPs/s > Higher Is Better 1 . 601 |====================================================================== 2 . 600 |====================================================================== NCNN 20201218 Target: Vulkan GPU - Model: mobilenet ms < Lower Is Better 1 . 12.78 |=================================================================== 2 . 12.92 |==================================================================== NCNN 20201218 Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2 ms < Lower Is Better 1 . 4.42 |===================================================================== 2 . 4.45 |===================================================================== NCNN 20201218 Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3 ms < Lower Is Better 1 . 4.11 |==================================================================== 2 . 4.18 |===================================================================== NCNN 20201218 Target: Vulkan GPU - Model: shufflenet-v2 ms < Lower Is Better 1 . 4.78 |===================================================================== 2 . 4.77 |===================================================================== NCNN 20201218 Target: Vulkan GPU - Model: mnasnet ms < Lower Is Better 1 . 4.12 |===================================================================== 2 . 4.04 |==================================================================== NCNN 20201218 Target: Vulkan GPU - Model: efficientnet-b0 ms < Lower Is Better 1 . 5.69 |===================================================================== 2 . 5.57 |==================================================================== NCNN 20201218 Target: Vulkan GPU - Model: blazeface ms < Lower Is Better 1 . 1.88 |===================================================================== 2 . 1.83 |=================================================================== NCNN 20201218 Target: Vulkan GPU - Model: googlenet ms < Lower Is Better 1 . 13.16 |==================================================================== 2 . 13.11 |==================================================================== NCNN 20201218 Target: Vulkan GPU - Model: vgg16 ms < Lower Is Better 1 . 55.24 |=================================================================== 2 . 56.10 |==================================================================== NCNN 20201218 Target: Vulkan GPU - Model: resnet18 ms < Lower Is Better 1 . 13.98 |==================================================================== 2 . 13.65 |================================================================== NCNN 20201218 Target: Vulkan GPU - Model: alexnet ms < Lower Is Better 1 . 11.06 |==================================================================== 2 . 11.08 |==================================================================== NCNN 20201218 Target: Vulkan GPU - Model: resnet50 ms < Lower Is Better 1 . 24.75 |==================================================================== 2 . 24.60 |==================================================================== NCNN 20201218 Target: Vulkan GPU - Model: yolov4-tiny ms < Lower Is Better 1 . 21.92 |==================================================================== 2 . 21.99 |==================================================================== NCNN 20201218 Target: Vulkan GPU - Model: squeezenet_ssd ms < Lower Is Better 1 . 14.93 |=================================================================== 2 . 15.19 |==================================================================== NCNN 20201218 Target: Vulkan GPU - Model: regnety_400m ms < Lower Is Better 1 . 16.75 |==================================================================== 2 . 16.74 |==================================================================== IndigoBench 4.4 Acceleration: OpenCL GPU - Scene: Bedroom M samples/s > Higher Is Better 1 . 20.87 |==================================================================== 2 . 20.87 |==================================================================== IndigoBench 4.4 Acceleration: OpenCL GPU - Scene: Supercar M samples/s > Higher Is Better 1 . 52.93 |==================================================================== 2 . 52.72 |==================================================================== Blender 2.92 Blend File: BMW27 - Compute: CUDA Seconds < Lower Is Better 1 . 18.35 |==================================================================== 2 . 18.36 |==================================================================== Blender 2.92 Blend File: Classroom - Compute: CUDA Seconds < Lower Is Better 1 . 51.42 |==================================================================== 2 . 51.42 |==================================================================== Blender 2.92 Blend File: Fishy Cat - Compute: CUDA Seconds < Lower Is Better 1 . 34.60 |==================================================================== 2 . 34.61 |==================================================================== Blender 2.92 Blend File: Barbershop - Compute: CUDA Seconds < Lower Is Better 1 . 373.50 |=================================================================== 2 . 374.21 |=================================================================== Blender 2.92 Blend File: BMW27 - Compute: NVIDIA OptiX Seconds < Lower Is Better 1 . 9.76 |===================================================================== 2 . 9.77 |===================================================================== Blender 2.92 Blend File: Classroom - Compute: NVIDIA OptiX Seconds < Lower Is Better 1 . 30.52 |==================================================================== 2 . 30.33 |==================================================================== Blender 2.92 Blend File: Fishy Cat - Compute: NVIDIA OptiX Seconds < Lower Is Better 1 . 20.43 |==================================================================== 2 . 20.42 |==================================================================== Blender 2.92 Blend File: Barbershop - Compute: NVIDIA OptiX Seconds < Lower Is Better 1 . 335.89 |=================================================================== 2 . 335.87 |=================================================================== Blender 2.92 Blend File: Pabellon Barcelona - Compute: CUDA Seconds < Lower Is Better 1 . 120.48 |=================================================================== 2 . 120.49 |=================================================================== Blender 2.92 Blend File: Pabellon Barcelona - Compute: NVIDIA OptiX Seconds < Lower Is Better 1 . 46.68 |==================================================================== 2 . 46.72 |==================================================================== MandelGPU 1.3pts1 OpenCL Device: GPU Samples/sec > Higher Is Better 1 . 466253751.5 |============================================================== 2 . 468356717.8 |============================================================== clpeak OpenCL Test: Integer Compute INT GIOPS > Higher Is Better 1 . 17922.35 |================================================================= 2 . 17951.73 |================================================================= clpeak OpenCL Test: Single-Precision Float GFLOPS > Higher Is Better 1 . 35173.57 |================================================================= 2 . 35204.07 |================================================================= clpeak OpenCL Test: Double-Precision Double GFLOPS > Higher Is Better 1 . 654.89 |=================================================================== 2 . 657.03 |=================================================================== clpeak OpenCL Test: Global Memory Bandwidth GBPS > Higher Is Better 1 . 810.12 |=================================================================== 2 . 813.37 |=================================================================== Chaos Group V-RAY 5 Mode: NVIDIA RTX GPU vrays > Higher Is Better 1 . 2601 |===================================================================== 2 . 2610 |===================================================================== Chaos Group V-RAY 5 Mode: NVIDIA CUDA GPU vpaths > Higher Is Better 1 . 1963 |===================================================================== 2 . 1963 |===================================================================== ViennaCL 1.7.1 Test: OpenCL BLAS - dGEMM-TT GFLOPs/s > Higher Is Better 2 . 605 |======================================================================