NVIDIA AMD Linux GPU Compute December 2018 NVIDIA and AMD GPU Linux compute benchmarks December 2018 by Michael Larabel for a future article.. GTX 1070: Processor: Intel Core i9-9900K @ 5.00GHz (8 Cores / 16 Threads), Motherboard: ASUS PRIME Z390-A (0602 BIOS), Chipset: Intel Cannon Lake PCH Shared SRAM, Memory: 16384MB, Disk: 2000GB SABRENT + Samsung SSD 970 EVO 250GB, Graphics: NVIDIA GeForce GTX 1070 8GB (1506/4006MHz), Audio: Realtek ALC1220, Monitor: Acer B286HK, Network: Intel Connection OS: Ubuntu 18.04, Kernel: 4.19.5-041905-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: NVIDIA 415.22, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 10.0.132, Vulkan: 1.1.84, Compiler: GCC 7.3.0 + CUDA 10.0, File-System: ext4, Screen Resolution: 3840x2160 GTX 1070 Ti: Processor: Intel Core i9-9900K @ 5.00GHz (8 Cores / 16 Threads), Motherboard: ASUS PRIME Z390-A (0602 BIOS), Chipset: Intel Cannon Lake PCH Shared SRAM, Memory: 16384MB, Disk: 2000GB SABRENT + Samsung SSD 970 EVO 250GB, Graphics: Zotac NVIDIA GeForce GTX 1070 Ti 8GB (1607/4006MHz), Audio: Realtek ALC1220, Monitor: Acer B286HK, Network: Intel Connection OS: Ubuntu 18.04, Kernel: 4.19.5-041905-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: NVIDIA 415.22, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 10.0.132, Vulkan: 1.1.84, Compiler: GCC 7.3.0 + CUDA 10.0, File-System: ext4, Screen Resolution: 3840x2160 GTX 1080: Processor: Intel Core i9-9900K @ 5.00GHz (8 Cores / 16 Threads), Motherboard: ASUS PRIME Z390-A (0602 BIOS), Chipset: Intel Cannon Lake PCH Shared SRAM, Memory: 16384MB, Disk: 2000GB SABRENT + Samsung SSD 970 EVO 250GB, Graphics: NVIDIA GeForce GTX 1080 8GB (1607/5005MHz), Audio: Realtek ALC1220, Monitor: Acer B286HK, Network: Intel Connection OS: Ubuntu 18.04, Kernel: 4.19.5-041905-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: NVIDIA 415.22, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 10.0.132, Vulkan: 1.1.84, Compiler: GCC 7.3.0 + CUDA 10.0, File-System: ext4, Screen Resolution: 3840x2160 GTX 1080 Ti: Processor: Intel Core i9-9900K @ 5.00GHz (8 Cores / 16 Threads), Motherboard: ASUS PRIME Z390-A (0602 BIOS), Chipset: Intel Cannon Lake PCH Shared SRAM, Memory: 16384MB, Disk: 2000GB SABRENT + Samsung SSD 970 EVO 250GB, Graphics: NVIDIA GeForce GTX 1080 Ti 11GB (1480/5508MHz), Audio: Realtek ALC1220, Monitor: Acer B286HK, Network: Intel Connection OS: Ubuntu 18.04, Kernel: 4.19.5-041905-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: NVIDIA 415.22, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 10.0.132, Vulkan: 1.1.84, Compiler: GCC 7.3.0 + CUDA 10.0, File-System: ext4, Screen Resolution: 3840x2160 RTX 2070: Processor: Intel Core i9-9900K @ 5.00GHz (8 Cores / 16 Threads), Motherboard: ASUS PRIME Z390-A (0602 BIOS), Chipset: Intel Cannon Lake PCH Shared SRAM, Memory: 16384MB, Disk: 2000GB SABRENT + Samsung SSD 970 EVO 250GB, Graphics: eVGA NVIDIA GeForce RTX 2070 8GB (1410/7000MHz), Audio: Realtek ALC1220, Monitor: Acer B286HK, Network: Intel Connection OS: Ubuntu 18.04, Kernel: 4.19.5-041905-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: NVIDIA 415.22, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 10.0.132, Vulkan: 1.1.84, Compiler: GCC 7.3.0 + CUDA 10.0, File-System: ext4, Screen Resolution: 3840x2160 RTX 2080: Processor: Intel Core i9-9900K @ 5.00GHz (8 Cores / 16 Threads), Motherboard: ASUS PRIME Z390-A (0602 BIOS), Chipset: Intel Cannon Lake PCH Shared SRAM, Memory: 16384MB, Disk: 2000GB SABRENT + Samsung SSD 970 EVO 250GB, Graphics: Zotac NVIDIA GeForce RTX 2080 8GB (1515/7000MHz), Audio: Realtek ALC1220, Monitor: Acer B286HK, Network: Intel Connection OS: Ubuntu 18.04, Kernel: 4.19.5-041905-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: NVIDIA 415.22, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 10.0.132, Vulkan: 1.1.84, Compiler: GCC 7.3.0 + CUDA 10.0, File-System: ext4, Screen Resolution: 3840x2160 RTX 2080 Ti: Processor: Intel Core i9-9900K @ 5.00GHz (8 Cores / 16 Threads), Motherboard: ASUS PRIME Z390-A (0602 BIOS), Chipset: Intel Cannon Lake PCH Shared SRAM, Memory: 16384MB, Disk: 2000GB SABRENT + Samsung SSD 970 EVO 250GB, Graphics: NVIDIA GeForce RTX 2080 Ti 11GB (1350/7000MHz), Audio: Realtek ALC1220, Monitor: Acer B286HK, Network: Intel Connection OS: Ubuntu 18.04, Kernel: 4.19.5-041905-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, Display Driver: NVIDIA 415.22, OpenGL: 4.6.0, OpenCL: OpenCL 1.2 CUDA 10.0.132, Vulkan: 1.1.84, Compiler: GCC 7.3.0 + CUDA 10.0, File-System: ext4, Screen Resolution: 3840x2160 R9 Fury: Processor: Intel Core i9-9900K @ 5.00GHz (8 Cores / 16 Threads), Motherboard: ASUS PRIME Z390-A (0602 BIOS), Chipset: Intel Cannon Lake PCH Shared SRAM, Memory: 16384MB, Disk: 2000GB SABRENT + Samsung SSD 970 EVO 250GB, Graphics: Sapphire AMD Radeon R9 FURY / NANO 4GB (1000/500MHz), Audio: Realtek ALC1220, Monitor: Acer B286HK, Network: Intel Connection OS: Ubuntu 18.04, Kernel: 4.19.5-041905-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, OpenGL: 4.5 Mesa 19.0.0-devel padoka PPA (LLVM 8.0.0), OpenCL: OpenCL 2.1 AMD-APP (2679.0), Vulkan: 1.1.70, Compiler: GCC 7.3.0, File-System: ext4, Screen Resolution: 3840x2160 RX Vega 56: Processor: Intel Core i9-9900K @ 5.00GHz (8 Cores / 16 Threads), Motherboard: ASUS PRIME Z390-A (0602 BIOS), Chipset: Intel Cannon Lake PCH Shared SRAM, Memory: 16384MB, Disk: 2000GB SABRENT + Samsung SSD 970 EVO 250GB, Graphics: AMD Radeon RX Vega 8GB (1590/800MHz), Audio: Realtek ALC1220, Monitor: Acer B286HK, Network: Intel Connection OS: Ubuntu 18.04, Kernel: 4.19.5-041905-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, OpenGL: 4.5 Mesa 19.0.0-devel padoka PPA (LLVM 8.0.0), OpenCL: OpenCL 2.1 AMD-APP (2679.0), Vulkan: 1.1.70, Compiler: GCC 7.3.0, File-System: ext4, Screen Resolution: 3840x2160 RX Vega 64: Processor: Intel Core i9-9900K @ 5.00GHz (8 Cores / 16 Threads), Motherboard: ASUS PRIME Z390-A (0602 BIOS), Chipset: Intel Cannon Lake PCH Shared SRAM, Memory: 16384MB, Disk: 2000GB SABRENT + Samsung SSD 970 EVO 250GB, Graphics: AMD Radeon RX Vega 8GB (1630/945MHz), Audio: Realtek ALC1220, Monitor: Acer B286HK, Network: Intel Connection OS: Ubuntu 18.04, Kernel: 4.19.5-041905-generic (x86_64), Desktop: GNOME Shell 3.28.3, Display Server: X Server 1.19.6, OpenGL: 4.5 Mesa 19.0.0-devel padoka PPA (LLVM 8.0.0), OpenCL: OpenCL 2.1 AMD-APP (2679.0), Vulkan: 1.1.70, Compiler: GCC 7.3.0, File-System: ext4, Screen Resolution: 3840x2160 SHOC Scalable HeterOgeneous Computing 2015-11-10 Target: OpenCL - Benchmark: FFT SP GFLOPS > Higher Is Better GTX 1070 .... 452 |================== GTX 1070 Ti . 497 |==================== GTX 1080 .... 575 |======================== GTX 1080 Ti . 972 |======================================== RTX 2070 .... 998 |========================================= RTX 2080 .... 1083 |============================================ RTX 2080 Ti . 1443 |=========================================================== R9 Fury ..... 846 |=================================== RX Vega 56 .. 926 |====================================== RX Vega 64 .. 1070 |============================================ SHOC Scalable HeterOgeneous Computing 2015-11-10 Target: OpenCL - Benchmark: MD5 Hash GHash/s > Higher Is Better GTX 1070 .... 10.72 |================= GTX 1070 Ti . 11.63 |=================== GTX 1080 .... 14.17 |======================= GTX 1080 Ti . 19.96 |================================ RTX 2070 .... 19.29 |=============================== RTX 2080 .... 24.37 |======================================= RTX 2080 Ti . 35.86 |========================================================== R9 Fury ..... 9.20 |=============== RX Vega 56 .. 14.03 |======================= RX Vega 64 .. 16.53 |=========================== SHOC Scalable HeterOgeneous Computing 2015-11-10 Target: OpenCL - Benchmark: Texture Read Bandwidth GB/s > Higher Is Better GTX 1070 .... 451 |======================= GTX 1070 Ti . 433 |======================= GTX 1080 .... 530 |============================ GTX 1080 Ti . 595 |=============================== RTX 2070 .... 1101 |========================================================= RTX 2080 .... 1119 |========================================================== RTX 2080 Ti . 1134 |=========================================================== R9 Fury ..... 250 |============= RX Vega 56 .. 384 |==================== RX Vega 64 .. 442 |======================= cl-mem 2017-01-13 Benchmark: Copy GB/s > Higher Is Better GTX 1070 .... 187 |========================= GTX 1070 Ti . 188 |========================= GTX 1080 .... 209 |============================ GTX 1080 Ti . 317 |========================================== RTX 2070 .... 330 |============================================ RTX 2080 .... 328 |=========================================== RTX 2080 Ti . 454 |============================================================ R9 Fury ..... 205 |=========================== RX Vega 56 .. 203 |=========================== RX Vega 64 .. 221 |============================= LuxMark 3.1 OpenCL Device: GPU - Scene: Luxball HDR Score > Higher Is Better GTX 1070 .... 17288 |======================= GTX 1070 Ti . 16886 |======================= GTX 1080 .... 13823 |=================== GTX 1080 Ti . 21562 |============================= RTX 2070 .... 30091 |========================================= RTX 2080 .... 29641 |======================================== RTX 2080 Ti . 42693 |========================================================== R9 Fury ..... 23448 |================================ RX Vega 56 .. 31179 |========================================== RX Vega 64 .. 32815 |============================================= clpeak OpenCL Test: Kernel Latency us < Lower Is Better GTX 1070 .... 3.74 |================================ GTX 1070 Ti . 3.66 |=============================== GTX 1080 .... 3.72 |=============================== GTX 1080 Ti . 3.80 |================================ RTX 2070 .... 3.52 |============================== RTX 2080 .... 3.48 |============================= RTX 2080 Ti . 3.57 |============================== R9 Fury ..... 5.68 |================================================ RX Vega 56 .. 6.98 |=========================================================== RX Vega 64 .. 6.94 |=========================================================== clpeak OpenCL Test: Integer Compute INT GIOPS > Higher Is Better GTX 1070 .... 1692 |======= GTX 1070 Ti . 2082 |======== GTX 1080 .... 2437 |========== GTX 1080 Ti . 3321 |============= RTX 2070 .... 8007 |================================ RTX 2080 .... 10059 |========================================= RTX 2080 Ti . 14385 |========================================================== R9 Fury ..... 1430 |====== RX Vega 56 .. 1991 |======== RX Vega 64 .. 2491 |========== SHOC Scalable HeterOgeneous Computing 2015-11-10 Target: OpenCL - Benchmark: Texture Read Bandwidth GB/s Per Watt > Higher Is Better GTX 1070 .... 3.35 |============================= GTX 1070 Ti . 3.53 |=============================== GTX 1080 .... 3.66 |================================ GTX 1080 Ti . 3.13 |=========================== RTX 2070 .... 6.76 |=========================================================== RTX 2080 .... 6.59 |========================================================== RTX 2080 Ti . 5.08 |============================================ R9 Fury ..... 1.45 |============= RX Vega 56 .. 1.94 |================= RX Vega 64 .. 1.98 |================= SHOC Scalable HeterOgeneous Computing 2015-11-10 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 83 AVG: 135 MAX: 155 GTX 1070 Ti . MIN: 47 AVG: 123 MAX: 142 GTX 1080 .... MIN: 90 AVG: 145 MAX: 165 GTX 1080 Ti . MIN: 121 AVG: 190 MAX: 217 RTX 2070 .... MIN: 54 AVG: 163 MAX: 199 RTX 2080 .... MIN: 47 AVG: 170 MAX: 203 RTX 2080 Ti . MIN: 48 AVG: 223 MAX: 283 R9 Fury ..... MIN: 114 AVG: 172 MAX: 221 RX Vega 56 .. MIN: 49 AVG: 198 MAX: 234 RX Vega 64 .. MIN: 52 AVG: 223 MAX: 250 cl-mem 2017-01-13 Benchmark: Copy GB/s Per Watt > Higher Is Better GTX 1070 .... 1.36 |===================== GTX 1070 Ti . 1.77 |=========================== GTX 1080 .... 1.40 |===================== GTX 1080 Ti . 1.77 |=========================== RTX 2070 .... 3.88 |=========================================================== RTX 2080 .... 3.53 |====================================================== RTX 2080 Ti . 2.52 |====================================== R9 Fury ..... 1.24 |=================== RX Vega 56 .. 1.03 |================ RX Vega 64 .. 1.02 |================ cl-mem 2017-01-13 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 83 AVG: 137 MAX: 149 GTX 1070 Ti . MIN: 46 AVG: 106 MAX: 138 GTX 1080 .... MIN: 115 AVG: 149 MAX: 162 GTX 1080 Ti . MIN: 121 AVG: 180 MAX: 209 RTX 2070 .... MIN: 43 AVG: 85 MAX: 108 RTX 2080 .... MIN: 46 AVG: 93 MAX: 122 RTX 2080 Ti . MIN: 47 AVG: 180 MAX: 249 R9 Fury ..... MIN: 90 AVG: 166 MAX: 211 RX Vega 56 .. MIN: 50 AVG: 197 MAX: 261 RX Vega 64 .. MIN: 52 AVG: 217 MAX: 319 LuxMark 3.1 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 83 AVG: 177 MAX: 181 GTX 1070 Ti . MIN: 147 AVG: 149 MAX: 149 GTX 1080 .... MIN: 91 AVG: 174 MAX: 176 GTX 1080 Ti . MIN: 123 AVG: 242 MAX: 248 RTX 2070 .... MIN: 78 AVG: 225 MAX: 233 RTX 2080 .... MIN: 102 AVG: 241 MAX: 247 RTX 2080 Ti . MIN: 111 AVG: 315 MAX: 328 R9 Fury ..... MIN: 90 AVG: 249 MAX: 257 RX Vega 56 .. MIN: 52 AVG: 251 MAX: 263 RX Vega 64 .. MIN: 53 AVG: 323 MAX: 338 GPU Temperature Monitor Phoronix Test Suite System Monitoring Celsius GTX 1070 .... MIN: 31.0 AVG: 61.1 MAX: 75.0 GTX 1070 Ti . MIN: 36.0 AVG: 46.9 MAX: 60.0 GTX 1080 .... MIN: 45.0 AVG: 63.1 MAX: 79.0 GTX 1080 Ti . MIN: 33.0 AVG: 66.5 MAX: 80.0 RTX 2070 .... MIN: 33.0 AVG: 54.6 MAX: 64.0 RTX 2080 .... MIN: 35.0 AVG: 66.5 MAX: 80.0 RTX 2080 Ti . MIN: 34.0 AVG: 60.3 MAX: 76.0 R9 Fury ..... MIN: 33.0 AVG: 67.0 MAX: 79.0 RX Vega 56 .. MIN: 31.0 AVG: 51.0 MAX: 75.0 RX Vega 64 .. MIN: 28.0 AVG: 53.0 MAX: 85.0 System Power Consumption Monitor Phoronix Test Suite System Monitoring Watts GTX 1070 .... MIN: 42 AVG: 144 MAX: 214 GTX 1070 Ti . MIN: 46 AVG: 128 MAX: 214 GTX 1080 .... MIN: 42 AVG: 155 MAX: 252 GTX 1080 Ti . MIN: 43 AVG: 201 MAX: 321 RTX 2070 .... MIN: 42 AVG: 146 MAX: 247 RTX 2080 .... MIN: 44 AVG: 175 MAX: 288 RTX 2080 Ti . MIN: 46 AVG: 208 MAX: 347 R9 Fury ..... MIN: 80 AVG: 160 MAX: 365 RX Vega 56 .. MIN: 49 AVG: 144 MAX: 284 RX Vega 64 .. MIN: 48 AVG: 169 MAX: 359 Chaos Group V-RAY 1.1.0 Mode: CUDA GPU Seconds < Lower Is Better GTX 1070 .... 90.43 |================================================== GTX 1070 Ti . 86.30 |================================================ GTX 1080 .... 102.07 |========================================================= GTX 1080 Ti . 66.41 |===================================== RTX 2070 .... 66.02 |===================================== RTX 2080 .... 72.07 |======================================== RTX 2080 Ti . 56.09 |=============================== Chaos Group V-RAY 1.1.0 GPU Temperature Monitor Celsius < Lower Is Better GTX 1070 .... MIN: 46.0 AVG: 62.6 MAX: 68.0 GTX 1070 Ti . MIN: 38.0 AVG: 47.8 MAX: 50.0 GTX 1080 .... MIN: 48.0 AVG: 62.9 MAX: 67.0 GTX 1080 Ti . MIN: 54.0 AVG: 68.1 MAX: 73.0 RTX 2070 .... MIN: 53.0 AVG: 60.7 MAX: 62.0 RTX 2080 .... MIN: 50.0 AVG: 68.2 MAX: 74.0 RTX 2080 Ti . MIN: 46.0 AVG: 60.0 MAX: 68.0 Chaos Group V-RAY 1.1.0 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 43 AVG: 151 MAX: 164 GTX 1070 Ti . MIN: 46 AVG: 131 MAX: 136 GTX 1080 .... MIN: 44 AVG: 153 MAX: 166 GTX 1080 Ti . MIN: 120 AVG: 219 MAX: 232 RTX 2070 .... MIN: 54 AVG: 186 MAX: 199 RTX 2080 .... MIN: 63 AVG: 192 MAX: 203 RTX 2080 Ti . MIN: 70 AVG: 243 MAX: 272 CUDA Mini-Nbody 2015-11-10 Test: Original (NBody^2)/s > Higher Is Better GTX 1070 .... 91.99 |============ GTX 1070 Ti . 112.00 |=============== GTX 1080 .... 111.00 |=============== GTX 1080 Ti . 186.00 |========================= RTX 2070 .... 244.00 |================================= RTX 2080 .... 304.00 |========================================= RTX 2080 Ti . 426.00 |========================================================= CUDA Mini-Nbody 2015-11-10 GPU Temperature Monitor Celsius < Lower Is Better GTX 1070 .... MIN: 61 AVG: 71 MAX: 75 GTX 1070 Ti . MIN: 45 AVG: 56 MAX: 60 GTX 1080 .... MIN: 64 AVG: 73 MAX: 79 GTX 1080 Ti . MIN: 65 AVG: 74 MAX: 80 RTX 2070 .... MIN: 55 AVG: 60 MAX: 64 RTX 2080 .... MIN: 65 AVG: 74 MAX: 78 RTX 2080 Ti . MIN: 58 AVG: 64 MAX: 67 CUDA Mini-Nbody 2015-11-10 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 44 AVG: 194 MAX: 211 GTX 1070 Ti . MIN: 47 AVG: 185 MAX: 214 GTX 1080 .... MIN: 44 AVG: 223 MAX: 252 GTX 1080 Ti . MIN: 51 AVG: 250 MAX: 313 RTX 2070 .... MIN: 54 AVG: 215 MAX: 246 RTX 2080 .... MIN: 49 AVG: 234 MAX: 284 RTX 2080 Ti . MIN: 49 AVG: 235 MAX: 336 NVIDIA GPU Cloud TensorFlow 18.09 Test: ResNet-50, FP16 Images Per Second > Higher Is Better GTX 1070 .... 160 |===================== GTX 1070 Ti . 175 |======================= GTX 1080 .... 193 |========================== GTX 1080 Ti . 271 |==================================== RTX 2080 .... 335 |============================================= RTX 2080 Ti . 449 |============================================================ RTX 2070 .... 309 |========================================= NVIDIA GPU Cloud TensorFlow 18.09 Test: ResNet-50, FP16 Images Per Second Per Watt > Higher Is Better GTX 1070 .... 0.92 |========================== GTX 1070 Ti . 1.14 |================================ GTX 1080 .... 1.02 |============================= GTX 1080 Ti . 1.22 |================================== RTX 2080 .... 1.86 |==================================================== RTX 2080 Ti . 2.10 |=========================================================== RTX 2070 .... 1.90 |===================================================== NVIDIA GPU Cloud TensorFlow 18.09 GPU Temperature Monitor Celsius < Lower Is Better GTX 1070 .... MIN: 58.0 AVG: 65.5 MAX: 70.0 GTX 1070 Ti . MIN: 46.0 AVG: 50.9 MAX: 55.0 GTX 1080 .... MIN: 61.0 AVG: 68.3 MAX: 75.0 GTX 1080 Ti . MIN: 61.0 AVG: 68.2 MAX: 75.0 RTX 2080 .... MIN: 60.0 AVG: 67.0 MAX: 75.0 RTX 2080 Ti . MIN: 55.0 AVG: 59.5 MAX: 65.0 RTX 2070 .... MIN: 32.0 AVG: 45.5 MAX: 58.0 NVIDIA GPU Cloud TensorFlow 18.09 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 43 AVG: 174 MAX: 213 GTX 1070 Ti . MIN: 47 AVG: 154 MAX: 192 GTX 1080 .... MIN: 45 AVG: 190 MAX: 249 GTX 1080 Ti . MIN: 55 AVG: 222 MAX: 320 RTX 2080 .... MIN: 47 AVG: 180 MAX: 285 RTX 2080 Ti . MIN: 56 AVG: 214 MAX: 342 RTX 2070 .... MIN: 71 AVG: 162 MAX: 247 NVIDIA GPU Cloud TensorFlow 18.09 Test: ResNet-50, FP32 Images Per Second > Higher Is Better GTX 1070 .... 125 |========================== GTX 1070 Ti . 133 |============================ GTX 1080 .... 143 |============================== GTX 1080 Ti . 210 |============================================ RTX 2080 .... 205 |=========================================== RTX 2080 Ti . 285 |============================================================ NVIDIA GPU Cloud TensorFlow 18.09 Test: ResNet-50, FP32 Images Per Second Per Watt > Higher Is Better GTX 1070 .... 0.73 |===================================== GTX 1070 Ti . 0.84 |=========================================== GTX 1080 .... 0.73 |===================================== GTX 1080 Ti . 0.89 |============================================== RTX 2080 .... 0.99 |=================================================== RTX 2080 Ti . 1.15 |=========================================================== NVIDIA GPU Cloud TensorFlow 18.09 GPU Temperature Monitor Celsius < Lower Is Better GTX 1070 .... MIN: 57 AVG: 66 MAX: 71 GTX 1070 Ti . MIN: 44 AVG: 51 MAX: 54 GTX 1080 .... MIN: 60 AVG: 69 MAX: 75 GTX 1080 Ti . MIN: 62 AVG: 70 MAX: 77 RTX 2080 .... MIN: 59 AVG: 70 MAX: 77 RTX 2080 Ti . MIN: 54 AVG: 61 MAX: 67 NVIDIA GPU Cloud TensorFlow 18.09 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 46 AVG: 172 MAX: 214 GTX 1070 Ti . MIN: 48 AVG: 157 MAX: 188 GTX 1080 .... MIN: 44 AVG: 196 MAX: 250 GTX 1080 Ti . MIN: 80 AVG: 235 MAX: 319 RTX 2080 .... MIN: 48 AVG: 207 MAX: 285 RTX 2080 Ti . MIN: 51 AVG: 248 MAX: 347 NVIDIA GPU Cloud TensorFlow 18.09 Test: AlexNet, FP16 Images Per Second > Higher Is Better GTX 1070 .... 1563 |===================== GTX 1070 Ti . 1683 |====================== GTX 1080 .... 1875 |========================= GTX 1080 Ti . 2674 |==================================== RTX 2080 .... 3153 |========================================== RTX 2080 Ti . 4432 |=========================================================== NVIDIA GPU Cloud TensorFlow 18.09 Test: AlexNet, FP16 Images Per Second Per Watt > Higher Is Better GTX 1070 .... 9.86 |======================= GTX 1070 Ti . 12.39 |============================= GTX 1080 .... 11.19 |========================== GTX 1080 Ti . 12.53 |============================= RTX 2080 .... 16.53 |====================================== RTX 2080 Ti . 24.95 |========================================================== NVIDIA GPU Cloud TensorFlow 18.09 GPU Temperature Monitor Celsius < Lower Is Better GTX 1070 .... MIN: 59.0 AVG: 64.1 MAX: 67.0 GTX 1070 Ti . MIN: 46.0 AVG: 49.9 MAX: 53.0 GTX 1080 .... MIN: 62.0 AVG: 66.5 MAX: 70.0 GTX 1080 Ti . MIN: 63.0 AVG: 67.2 MAX: 70.0 RTX 2080 .... MIN: 63.0 AVG: 67.6 MAX: 72.0 RTX 2080 Ti . MIN: 57.0 AVG: 60.4 MAX: 64.0 NVIDIA GPU Cloud TensorFlow 18.09 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 69 AVG: 158 MAX: 192 GTX 1070 Ti . MIN: 47 AVG: 136 MAX: 175 GTX 1080 .... MIN: 44 AVG: 168 MAX: 235 GTX 1080 Ti . MIN: 51 AVG: 213 MAX: 303 RTX 2080 .... MIN: 47 AVG: 191 MAX: 270 RTX 2080 Ti . MIN: 49 AVG: 178 MAX: 333 NVIDIA GPU Cloud TensorFlow 18.09 Test: AlexNet, FP32 Images Per Second > Higher Is Better GTX 1070 .... 1516 |=========================== GTX 1070 Ti . 1591 |============================ GTX 1080 .... 1764 |=============================== GTX 1080 Ti . 2539 |============================================= RTX 2070 .... 2174 |====================================== RTX 2080 .... 2419 |=========================================== RTX 2080 Ti . 3344 |=========================================================== NVIDIA GPU Cloud TensorFlow 18.09 Test: AlexNet, FP32 Images Per Second Per Watt > Higher Is Better GTX 1070 .... 9.61 |=========================================== GTX 1070 Ti . 10.99 |================================================= GTX 1080 .... 9.90 |============================================ GTX 1080 Ti . 12.19 |====================================================== RTX 2070 .... 11.72 |==================================================== RTX 2080 .... 11.44 |=================================================== RTX 2080 Ti . 12.99 |========================================================== NVIDIA GPU Cloud TensorFlow 18.09 GPU Temperature Monitor Celsius < Lower Is Better GTX 1070 .... MIN: 61.0 AVG: 64.2 MAX: 68.0 GTX 1070 Ti . MIN: 44.0 AVG: 49.1 MAX: 52.0 GTX 1080 .... MIN: 63.0 AVG: 66.8 MAX: 71.0 GTX 1080 Ti . MIN: 64.0 AVG: 67.6 MAX: 71.0 RTX 2070 .... MIN: 46.0 AVG: 55.9 MAX: 61.0 RTX 2080 .... MIN: 59.0 AVG: 69.8 MAX: 76.0 RTX 2080 Ti . MIN: 53.0 AVG: 61.4 MAX: 67.0 NVIDIA GPU Cloud TensorFlow 18.09 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 87 AVG: 158 MAX: 202 GTX 1070 Ti . MIN: 48 AVG: 145 MAX: 186 GTX 1080 .... MIN: 101 AVG: 178 MAX: 229 GTX 1080 Ti . MIN: 126 AVG: 208 MAX: 307 RTX 2070 .... MIN: 64 AVG: 185 MAX: 244 RTX 2080 .... MIN: 48 AVG: 211 MAX: 284 RTX 2080 Ti . MIN: 49 AVG: 257 MAX: 339 NVIDIA GPU Cloud TensorFlow 18.09 Test: Googlenet, FP16 Images Per Second > Higher Is Better GTX 1070 .... 375 |====================== GTX 1070 Ti . 413 |======================== GTX 1080 .... 458 |=========================== GTX 1080 Ti . 629 |===================================== RTX 2080 .... 738 |=========================================== RTX 2080 Ti . 1015 |=========================================================== NVIDIA GPU Cloud TensorFlow 18.09 Test: Googlenet, FP16 Images Per Second Per Watt > Higher Is Better GTX 1070 .... 2.06 |============================== GTX 1070 Ti . 2.53 |===================================== GTX 1080 .... 2.13 |=============================== GTX 1080 Ti . 2.45 |==================================== RTX 2080 .... 3.21 |=============================================== RTX 2080 Ti . 4.07 |=========================================================== NVIDIA GPU Cloud TensorFlow 18.09 GPU Temperature Monitor Celsius < Lower Is Better GTX 1070 .... MIN: 56 AVG: 69 MAX: 74 GTX 1070 Ti . MIN: 43 AVG: 53 MAX: 56 GTX 1080 .... MIN: 58 AVG: 72 MAX: 79 GTX 1080 Ti . MIN: 60 AVG: 73 MAX: 80 RTX 2080 .... MIN: 60 AVG: 73 MAX: 80 RTX 2080 Ti . MIN: 55 AVG: 65 MAX: 71 NVIDIA GPU Cloud TensorFlow 18.09 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 43 AVG: 182 MAX: 211 GTX 1070 Ti . MIN: 47 AVG: 164 MAX: 188 GTX 1080 .... MIN: 76 AVG: 215 MAX: 247 GTX 1080 Ti . MIN: 64 AVG: 257 MAX: 319 RTX 2080 .... MIN: 48 AVG: 230 MAX: 285 RTX 2080 Ti . MIN: 50 AVG: 249 MAX: 343 NVIDIA GPU Cloud TensorFlow 18.09 Test: Inception v4, FP16 Images Per Second > Higher Is Better GTX 1070 .... 44.77 |=================== GTX 1070 Ti . 49.23 |===================== GTX 1080 .... 55.03 |======================= GTX 1080 Ti . 74.63 |=============================== RTX 2080 .... 102.77 |=========================================== RTX 2080 Ti . 135.90 |========================================================= NVIDIA GPU Cloud TensorFlow 18.09 Test: Inception v4, FP16 Images Per Second Per Watt > Higher Is Better GTX 1070 .... 0.28 |====================== GTX 1070 Ti . 0.34 |=========================== GTX 1080 .... 0.30 |======================== GTX 1080 Ti . 0.35 |============================ RTX 2080 .... 0.58 |============================================== RTX 2080 Ti . 0.74 |=========================================================== NVIDIA GPU Cloud TensorFlow 18.09 GPU Temperature Monitor Celsius < Lower Is Better GTX 1070 .... MIN: 55 AVG: 64 MAX: 72 GTX 1070 Ti . MIN: 43 AVG: 50 MAX: 55 GTX 1080 .... MIN: 51 AVG: 64 MAX: 76 GTX 1080 Ti . MIN: 56 AVG: 66 MAX: 77 RTX 2080 .... MIN: 57 AVG: 65 MAX: 74 RTX 2080 Ti . MIN: 53 AVG: 58 MAX: 65 NVIDIA GPU Cloud TensorFlow 18.09 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 44 AVG: 158 MAX: 210 GTX 1070 Ti . MIN: 47 AVG: 146 MAX: 189 GTX 1080 .... MIN: 71 AVG: 186 MAX: 250 GTX 1080 Ti . MIN: 52 AVG: 212 MAX: 321 RTX 2080 .... MIN: 48 AVG: 177 MAX: 285 RTX 2080 Ti . MIN: 50 AVG: 183 MAX: 341 NVIDIA GPU Cloud TensorFlow 18.09 Test: VGG-16, FP16 Images Per Second > Higher Is Better GTX 1070 .... 76.93 |===================== GTX 1070 Ti . 84.43 |======================= GTX 1080 .... 93.13 |========================== GTX 1080 Ti . 131.37 |==================================== RTX 2080 .... 153.33 |========================================== RTX 2080 Ti . 206.97 |========================================================= NVIDIA GPU Cloud TensorFlow 18.09 Test: VGG-16, FP16 Images Per Second Per Watt > Higher Is Better GTX 1070 .... 0.46 |============================= GTX 1070 Ti . 0.53 |================================= GTX 1080 .... 0.48 |============================== GTX 1080 Ti . 0.54 |================================== RTX 2080 .... 0.69 |=========================================== RTX 2080 Ti . 0.94 |=========================================================== NVIDIA GPU Cloud TensorFlow 18.09 GPU Temperature Monitor Celsius < Lower Is Better GTX 1070 .... MIN: 59 AVG: 66 MAX: 70 GTX 1070 Ti . MIN: 45 AVG: 52 MAX: 55 GTX 1080 .... MIN: 50 AVG: 66 MAX: 74 GTX 1080 Ti . MIN: 61 AVG: 70 MAX: 76 RTX 2080 .... MIN: 59 AVG: 70 MAX: 77 RTX 2080 Ti . MIN: 54 AVG: 62 MAX: 68 NVIDIA GPU Cloud TensorFlow 18.09 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 44 AVG: 169 MAX: 202 GTX 1070 Ti . MIN: 47 AVG: 160 MAX: 192 GTX 1080 .... MIN: 91 AVG: 195 MAX: 240 GTX 1080 Ti . MIN: 52 AVG: 243 MAX: 313 RTX 2080 .... MIN: 48 AVG: 224 MAX: 283 RTX 2080 Ti . MIN: 51 AVG: 220 MAX: 332 NVIDIA GPU Cloud TensorFlow 18.09 Test: VGG-16, FP32 Images Per Second > Higher Is Better GTX 1070 .... 71.57 |=========================== GTX 1070 Ti . 76.40 |============================= GTX 1080 .... 82.90 |=============================== GTX 1080 Ti . 119.50 |============================================= RTX 2080 .... 108.57 |========================================= RTX 2080 Ti . 152.40 |========================================================= NVIDIA GPU Cloud TensorFlow 18.09 Test: VGG-16, FP32 Images Per Second Per Watt > Higher Is Better GTX 1070 .... 0.42 |============================================ GTX 1070 Ti . 0.47 |================================================== GTX 1080 .... 0.42 |============================================ GTX 1080 Ti . 0.49 |==================================================== RTX 2080 .... 0.45 |=============================================== RTX 2080 Ti . 0.56 |=========================================================== NVIDIA GPU Cloud TensorFlow 18.09 GPU Temperature Monitor Celsius < Lower Is Better GTX 1070 .... MIN: 58 AVG: 66 MAX: 70 GTX 1070 Ti . MIN: 46 AVG: 52 MAX: 55 GTX 1080 .... MIN: 60 AVG: 69 MAX: 74 GTX 1080 Ti . MIN: 61 AVG: 71 MAX: 77 RTX 2080 .... MIN: 61 AVG: 74 MAX: 80 RTX 2080 Ti . MIN: 56 AVG: 66 MAX: 71 NVIDIA GPU Cloud TensorFlow 18.09 System Power Consumption Monitor Watts < Lower Is Better GTX 1070 .... MIN: 54 AVG: 171 MAX: 204 GTX 1070 Ti . MIN: 59 AVG: 162 MAX: 191 GTX 1080 .... MIN: 46 AVG: 198 MAX: 236 GTX 1080 Ti . MIN: 51 AVG: 242 MAX: 310 RTX 2080 .... MIN: 56 AVG: 239 MAX: 288 RTX 2080 Ti . MIN: 50 AVG: 270 MAX: 342 SHOC Scalable HeterOgeneous Computing 2015-11-10 Performance / Cost - Target: OpenCL - Benchmark: FFT SP GFLOPS Per Dollar > Higher Is Better GTX 1070 .... 1.13 |============================== GTX 1070 Ti . 1.11 |============================= GTX 1080 .... 1.05 |=========================== GTX 1080 Ti . 1.39 |==================================== RTX 2070 .... 1.67 |============================================ RTX 2080 .... 1.36 |==================================== RTX 2080 Ti . 1.20 |=============================== RX Vega 56 .. 2.26 |=========================================================== RX Vega 64 .. 2.25 |=========================================================== SHOC Scalable HeterOgeneous Computing 2015-11-10 Performance / Cost - Target: OpenCL - Benchmark: MD5 Hash GHash/s Per Dollar > Higher Is Better GTX 1070 .... 0.03 |=========================================================== GTX 1070 Ti . 0.03 |=========================================================== GTX 1080 .... 0.03 |=========================================================== GTX 1080 Ti . 0.03 |=========================================================== RTX 2070 .... 0.03 |=========================================================== RTX 2080 .... 0.03 |=========================================================== RTX 2080 Ti . 0.03 |=========================================================== RX Vega 56 .. 0.03 |=========================================================== RX Vega 64 .. 0.03 |=========================================================== SHOC Scalable HeterOgeneous Computing 2015-11-10 Performance / Cost - Target: OpenCL - Benchmark: Texture Read Bandwidth GB/s Per Dollar > Higher Is Better GTX 1070 .... 1.13 |==================================== GTX 1070 Ti . 0.96 |=============================== GTX 1080 .... 0.97 |=============================== GTX 1080 Ti . 0.85 |=========================== RTX 2070 .... 1.84 |=========================================================== RTX 2080 .... 1.40 |============================================= RTX 2080 Ti . 0.95 |============================== RX Vega 56 .. 0.94 |============================== RX Vega 64 .. 0.93 |============================== cl-mem 2017-01-13 Performance / Cost - Benchmark: Copy GB/s Per Dollar > Higher Is Better GTX 1070 .... 0.47 |================================================== GTX 1070 Ti . 0.42 |============================================= GTX 1080 .... 0.38 |========================================= GTX 1080 Ti . 0.45 |================================================ RTX 2070 .... 0.55 |=========================================================== RTX 2080 .... 0.41 |============================================ RTX 2080 Ti . 0.38 |========================================= RX Vega 56 .. 0.50 |====================================================== RX Vega 64 .. 0.47 |================================================== LuxMark 3.1 Performance / Cost - OpenCL Device: GPU - Scene: Luxball HDR Score Per Dollar > Higher Is Better GTX 1070 .... 43.33 |================================= GTX 1070 Ti . 37.61 |============================= GTX 1080 .... 25.18 |=================== GTX 1080 Ti . 30.85 |======================= RTX 2070 .... 50.24 |====================================== RTX 2080 .... 37.14 |============================ RTX 2080 Ti . 35.61 |=========================== RX Vega 56 .. 76.23 |========================================================== RX Vega 64 .. 69.08 |===================================================== clpeak Performance / Cost - OpenCL Test: Kernel Latency us x Dollar < Lower Is Better GTX 1070 .... 1492.26 |==================== GTX 1070 Ti . 1643.34 |===================== GTX 1080 .... 2042.28 |=========================== GTX 1080 Ti . 2656.20 |=================================== RTX 2070 .... 2108.48 |============================ RTX 2080 .... 2777.04 |==================================== RTX 2080 Ti . 4280.43 |======================================================== RX Vega 56 .. 2854.82 |===================================== RX Vega 64 .. 3296.50 |=========================================== clpeak Performance / Cost - OpenCL Test: Integer Compute INT GIOPS Per Dollar > Higher Is Better GTX 1070 .... 4.24 |================== GTX 1070 Ti . 4.64 |==================== GTX 1080 .... 4.44 |=================== GTX 1080 Ti . 4.75 |===================== RTX 2070 .... 13.37 |========================================================== RTX 2080 .... 12.61 |======================================================= RTX 2080 Ti . 12.00 |==================================================== RX Vega 56 .. 4.87 |===================== RX Vega 64 .. 5.24 |=======================