EPYC vs. Xeon vs. NVIDIA AMD GPUs POCL OpenCL Tests for a future article. AMD EPYC 7601: Processor: AMD EPYC 7601 32-Core @ 2.20GHz (32 Cores / 64 Threads), Motherboard: TYAN B8026T70AE24HR, Chipset: AMD Device 1450, Memory: 129024MB, Disk: 280GB INTEL SSDPE21D280GA, Graphics: ASPEED ASPEED Family, Monitor: VE228, Network: Broadcom Limited NetXtreme BCM5720 Gigabit PCIe OS: Ubuntu 17.10, Kernel: 4.15.0-999-generic (x86_64) 20171205, Desktop: GNOME Shell 3.26.1, Display Driver: modesetting 1.19.5, OpenGL: 3.3 Mesa 17.2.2 (LLVM 5.0 128 bits), OpenCL: OpenCL 1.2 pocl 1.0 LLVM 5.0.0, Compiler: GCC 7.2.0 + Clang 5.0.0-3 + LLVM 5.0.0, File-System: ext4, Screen Resolution: 1920x1080 2 x Intel Xeon Gold 6138: Processor: 2 x Intel Xeon Gold 6138 @ 3.70GHz (40 Cores / 80 Threads), Motherboard: TYAN S7106, Chipset: Intel Sky Lake-E DMI3 Registers, Memory: 96256MB, Disk: 256GB Samsung SSD 850 + 2000GB Seagate ST2000DM006-2DM1 + 2 x 120GB TOSHIBA-TR150, Graphics: ASPEED ASPEED Family, Monitor: VE228, Network: Intel I210 Gigabit Connection OS: Ubuntu 18.04, Kernel: 4.15.0-999-generic (x86_64) 20171205, Desktop: GNOME Shell 3.26.2, Display Driver: modesetting 1.19.5, OpenGL: 3.3 Mesa 17.2.2 (LLVM 5.0 256 bits), OpenCL: OpenCL 1.2 pocl 1.1-pre LLVM 5.0.0, Compiler: GCC 7.2.1 20171205 + Clang 5.0.0-4 + LLVM 5.0.0, File-System: ext4, Screen Resolution: 1920x1080 Radeon RX 580: Processor: Intel Core i7-8700K @ 4.70GHz (6 Cores / 12 Threads), Motherboard: ASUS PRIME Z370-A, Chipset: Intel Device 3ec2, Memory: 16384MB, Disk: 525GB Crucial_CT525MX3 + Samsung SSD 950 PRO 256GB, Graphics: MSI AMD Radeon RX 580 8192MB, Audio: Realtek Generic, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 17.04, Kernel: 4.10.0-38-generic (x86_64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.3, Display Driver: modesetting 1.19.3, OpenGL: 4.5.13496, OpenCL: OpenCL 2.0 AMD-APP (2482.3), Vulkan: 1.0.42, Compiler: GCC 6.3.0 20170406, File-System: ext4, Screen Resolution: 3840x2160 Radeon R9 Fury: Processor: Intel Core i7-8700K @ 4.70GHz (6 Cores / 12 Threads), Motherboard: ASUS PRIME Z370-A, Chipset: Intel Device 3ec2, Memory: 16384MB, Disk: 525GB Crucial_CT525MX3 + Samsung SSD 950 PRO 256GB, Graphics: Sapphire AMD Radeon 4096MB, Audio: Realtek Generic, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 17.04, Kernel: 4.10.0-38-generic (x86_64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.3, Display Driver: modesetting 1.19.3, OpenGL: 4.5.13496, OpenCL: OpenCL 2.0 AMD-APP (2482.3), Vulkan: 1.0.42, Compiler: GCC 6.3.0 20170406, File-System: ext4, Screen Resolution: 3840x2160 Radeon RX Vega 56: Processor: Intel Core i7-8700K @ 4.70GHz (6 Cores / 12 Threads), Motherboard: ASUS PRIME Z370-A, Chipset: Intel Device 3ec2, Memory: 16384MB, Disk: 525GB Crucial_CT525MX3 + Samsung SSD 950 PRO 256GB, Graphics: AMD Radeon RX Vega 8176MB, Audio: Realtek Generic, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 17.04, Kernel: 4.10.0-38-generic (x86_64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.3, Display Driver: amdgpu 1.3.99, OpenGL: 4.5.13496, OpenCL: OpenCL 2.0 AMD-APP (2482.3), Vulkan: 1.0.42, Compiler: GCC 6.3.0 20170406, File-System: ext4, Screen Resolution: 3840x2160 Radeon RX Vega 64: Processor: Intel Core i7-8700K @ 4.70GHz (6 Cores / 12 Threads), Motherboard: ASUS PRIME Z370-A, Chipset: Intel Device 3ec2, Memory: 16384MB, Disk: 525GB Crucial_CT525MX3 + Samsung SSD 950 PRO 256GB, Graphics: AMD Radeon RX Vega 8176MB, Audio: Realtek Generic, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 17.04, Kernel: 4.10.0-38-generic (x86_64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.3, Display Driver: amdgpu 1.3.99, OpenGL: 4.5.13496, OpenCL: OpenCL 2.0 AMD-APP (2482.3), Vulkan: 1.0.42, Compiler: GCC 6.3.0 20170406, File-System: ext4, Screen Resolution: 3840x2160 GeForce GTX 980 Ti: Processor: Intel Core i7-8700K @ 4.70GHz (6 Cores / 12 Threads), Motherboard: ASUS PRIME Z370-A, Chipset: Intel Device 3ec2, Memory: 16384MB, Disk: 525GB Crucial_CT525MX3 + Samsung SSD 950 PRO 256GB, Graphics: NVIDIA GeForce GTX 980 Ti 6144MB (999/3505MHz), Audio: Realtek Generic, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 17.04, Kernel: 4.10.0-38-generic (x86_64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.3, Display Driver: NVIDIA 384.98, OpenGL: 4.5.0, Vulkan: 1.0.42, Compiler: GCC 6.3.0 20170406, File-System: ext4, Screen Resolution: 3840x2160 GeForce GTX 1050: Processor: Intel Core i7-8700K @ 4.70GHz (6 Cores / 12 Threads), Motherboard: ASUS PRIME Z370-A, Chipset: Intel Device 3ec2, Memory: 16384MB, Disk: 525GB Crucial_CT525MX3 + Samsung SSD 950 PRO 256GB, Graphics: Zotac NVIDIA GeForce GTX 1050 2048MB (1354/3504MHz), Audio: Realtek Generic, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 17.04, Kernel: 4.10.0-38-generic (x86_64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.3, Display Driver: NVIDIA 384.98, OpenGL: 4.5.0, Vulkan: 1.0.42, Compiler: GCC 6.3.0 20170406, File-System: ext4, Screen Resolution: 3840x2160 GeForce GTX 1050 Ti: Processor: Intel Core i7-8700K @ 4.70GHz (6 Cores / 12 Threads), Motherboard: ASUS PRIME Z370-A, Chipset: Intel Device 3ec2, Memory: 16384MB, Disk: 525GB Crucial_CT525MX3 + Samsung SSD 950 PRO 256GB, Graphics: eVGA NVIDIA GeForce GTX 1050 Ti 4096MB (1354/3504MHz), Audio: Realtek Generic, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 17.04, Kernel: 4.10.0-38-generic (x86_64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.3, Display Driver: NVIDIA 384.98, OpenGL: 4.5.0, Vulkan: 1.0.42, Compiler: GCC 6.3.0 20170406, File-System: ext4, Screen Resolution: 3840x2160 GeForce GTX 1060: Processor: Intel Core i7-8700K @ 4.70GHz (6 Cores / 12 Threads), Motherboard: ASUS PRIME Z370-A, Chipset: Intel Device 3ec2, Memory: 16384MB, Disk: 525GB Crucial_CT525MX3 + Samsung SSD 950 PRO 256GB, Graphics: NVIDIA GeForce GTX 1060 6GB 6144MB (1506/4006MHz), Audio: Realtek Generic, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 17.04, Kernel: 4.10.0-38-generic (x86_64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.3, Display Driver: NVIDIA 384.98, OpenGL: 4.5.0, Vulkan: 1.0.42, Compiler: GCC 6.3.0 20170406, File-System: ext4, Screen Resolution: 3840x2160 GeForce GTX 1070: Processor: Intel Core i7-8700K @ 4.70GHz (6 Cores / 12 Threads), Motherboard: ASUS PRIME Z370-A, Chipset: Intel Device 3ec2, Memory: 16384MB, Disk: 525GB Crucial_CT525MX3 + Samsung SSD 950 PRO 256GB, Graphics: NVIDIA GeForce GTX 1070 8192MB (1506/4006MHz), Audio: Realtek Generic, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 17.04, Kernel: 4.10.0-38-generic (x86_64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.3, Display Driver: NVIDIA 384.98, OpenGL: 4.5.0, Vulkan: 1.0.42, Compiler: GCC 6.3.0 20170406, File-System: ext4, Screen Resolution: 3840x2160 GeForce GTX 1070 Ti: Processor: Intel Core i7-8700K @ 4.70GHz (6 Cores / 12 Threads), Motherboard: ASUS PRIME Z370-A, Chipset: Intel Device 3ec2, Memory: 16384MB, Disk: 525GB Crucial_CT525MX3 + Samsung SSD 950 PRO 256GB, Graphics: Zotac NVIDIA GeForce GTX 1070 Ti 8192MB (1480/4006MHz), Audio: Realtek Generic, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 17.04, Kernel: 4.10.0-38-generic (x86_64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.3, Display Driver: NVIDIA 384.98, OpenGL: 4.5.0, Vulkan: 1.0.42, Compiler: GCC 6.3.0 20170406, File-System: ext4, Screen Resolution: 3840x2160 GeForce GTX 1080: Processor: Intel Core i7-8700K @ 4.70GHz (6 Cores / 12 Threads), Motherboard: ASUS PRIME Z370-A, Chipset: Intel Device 3ec2, Memory: 16384MB, Disk: 525GB Crucial_CT525MX3 + Samsung SSD 950 PRO 256GB, Graphics: NVIDIA GeForce GTX 1080 8192MB (1607/5005MHz), Audio: Realtek Generic, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 17.04, Kernel: 4.10.0-38-generic (x86_64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.3, Display Driver: NVIDIA 384.98, OpenGL: 4.5.0, Vulkan: 1.0.42, Compiler: GCC 6.3.0 20170406, File-System: ext4, Screen Resolution: 3840x2160 GeForce GTX 1080 Ti: Processor: Intel Core i7-8700K @ 4.70GHz (6 Cores / 12 Threads), Motherboard: ASUS PRIME Z370-A, Chipset: Intel Device 3ec2, Memory: 16384MB, Disk: 525GB Crucial_CT525MX3 + Samsung SSD 950 PRO 256GB, Graphics: NVIDIA GeForce GTX 1080 Ti 11264MB (1480/5508MHz), Audio: Realtek Generic, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 17.04, Kernel: 4.10.0-38-generic (x86_64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.3, Display Driver: NVIDIA 384.98, OpenGL: 4.5.0, Vulkan: 1.0.42, Compiler: GCC 6.3.0 20170406, File-System: ext4, Screen Resolution: 3840x2160 SHOC Scalable HeterOgeneous Computing 2015-11-10 Target: OpenCL - Benchmark: Triad GB/s > Higher Is Better AMD EPYC 7601 ............ 6.95 |========================= 2 x Intel Xeon Gold 6138 . 3.26 |============ Radeon RX 580 ............ 12.02 |=========================================== Radeon R9 Fury ........... 11.39 |========================================= Radeon RX Vega 56 ........ 12.47 |============================================= Radeon RX Vega 64 ........ 11.94 |=========================================== GeForce GTX 980 Ti ....... 12.26 |============================================ GeForce GTX 1050 ......... 11.44 |========================================= GeForce GTX 1050 Ti ...... 11.46 |========================================= GeForce GTX 1060 ......... 11.94 |=========================================== GeForce GTX 1070 ......... 12.18 |============================================ GeForce GTX 1070 Ti ...... 12.18 |============================================ GeForce GTX 1080 ......... 12.28 |============================================ GeForce GTX 1080 Ti ...... 12.46 |============================================= SHOC Scalable HeterOgeneous Computing 2015-11-10 Target: OpenCL - Benchmark: FFT SP GFLOPS > Higher Is Better AMD EPYC 7601 ............ 13.49 |= 2 x Intel Xeon Gold 6138 . 10.78 | Radeon RX 580 ............ 521.55 |===================== Radeon R9 Fury ........... 793.70 |================================ Radeon RX Vega 56 ........ 917.73 |===================================== Radeon RX Vega 64 ........ 1072.78 |=========================================== GeForce GTX 980 Ti ....... 652.04 |========================== GeForce GTX 1050 ......... 197.20 |======== GeForce GTX 1050 Ti ...... 197.93 |======== GeForce GTX 1060 ......... 333.30 |============= GeForce GTX 1070 ......... 481.09 |=================== GeForce GTX 1070 Ti ...... 514.97 |===================== GeForce GTX 1080 ......... 582.56 |======================= GeForce GTX 1080 Ti ...... 990.47 |======================================== SHOC Scalable HeterOgeneous Computing 2015-11-10 Target: OpenCL - Benchmark: MD5 Hash GHash/s > Higher Is Better AMD EPYC 7601 ............ 0.40 |= 2 x Intel Xeon Gold 6138 . 0.48 |= Radeon RX 580 ............ 7.73 |================= Radeon R9 Fury ........... 8.84 |==================== Radeon RX Vega 56 ........ 14.23 |================================ Radeon RX Vega 64 ........ 16.99 |====================================== GeForce GTX 980 Ti ....... 9.28 |===================== GeForce GTX 1050 ......... 3.24 |======= GeForce GTX 1050 Ti ...... 4.11 |========= GeForce GTX 1060 ......... 7.31 |================= GeForce GTX 1070 ......... 10.60 |======================== GeForce GTX 1070 Ti ...... 13.80 |=============================== GeForce GTX 1080 ......... 14.30 |================================ GeForce GTX 1080 Ti ...... 19.90 |============================================= SHOC Scalable HeterOgeneous Computing 2015-11-10 Target: OpenCL - Benchmark: Max SP Flops GFLOPS > Higher Is Better AMD EPYC 7601 ............ 703.34 |== 2 x Intel Xeon Gold 6138 . 1031.48 |=== Radeon RX 580 ............ 6261.97 |==================== Radeon R9 Fury ........... 7141.33 |====================== Radeon RX Vega 56 ........ 10702.80 |================================== Radeon RX Vega 64 ........ 12923.60 |======================================== GeForce GTX 980 Ti ....... 6250.35 |==================== GeForce GTX 1050 ......... 2123.34 |======= GeForce GTX 1050 Ti ...... 2702.95 |======== GeForce GTX 1060 ......... 4854.00 |=============== GeForce GTX 1070 ......... 7256.76 |======================= GeForce GTX 1070 Ti ...... 9127.64 |============================= GeForce GTX 1080 ......... 9593.38 |============================== GeForce GTX 1080 Ti ...... 13411.90 |========================================== SHOC Scalable HeterOgeneous Computing 2015-11-10 Target: OpenCL - Benchmark: Texture Read Bandwidth GB/s > Higher Is Better AMD EPYC 7601 ............ 41.84 |=== 2 x Intel Xeon Gold 6138 . 40.35 |=== Radeon RX 580 ............ 216.56 |================ Radeon R9 Fury ........... 249.34 |================== Radeon RX Vega 56 ........ 377.52 |============================ Radeon RX Vega 64 ........ 431.29 |================================ GeForce GTX 980 Ti ....... 351.98 |========================== GeForce GTX 1050 ......... 267.71 |==================== GeForce GTX 1050 Ti ...... 303.88 |====================== GeForce GTX 1060 ......... 377.95 |============================ GeForce GTX 1070 ......... 451.38 |================================= GeForce GTX 1070 Ti ...... 503.41 |===================================== GeForce GTX 1080 ......... 524.60 |======================================= GeForce GTX 1080 Ti ...... 598.58 |============================================ ViennaCL 1.4.2 OpenCL LU Factorization GFLOPS > Higher Is Better AMD EPYC 7601 ............ 14.20 |========== 2 x Intel Xeon Gold 6138 . 20.32 |============== Radeon RX 580 ............ 14.13 |========== Radeon R9 Fury ........... 24.55 |================= Radeon RX Vega 56 ........ 41.47 |============================= Radeon RX Vega 64 ........ 42.58 |============================= GeForce GTX 980 Ti ....... 59.24 |========================================= GeForce GTX 1050 ......... 43.41 |============================== GeForce GTX 1050 Ti ...... 47.55 |================================= GeForce GTX 1060 ......... 56.09 |======================================= GeForce GTX 1070 ......... 59.66 |========================================= GeForce GTX 1070 Ti ...... 63.24 |============================================ GeForce GTX 1080 ......... 63.58 |============================================ GeForce GTX 1080 Ti ...... 65.24 |============================================= cl-mem 2017-01-13 Benchmark: Copy GB/s > Higher Is Better AMD EPYC 7601 ............ 6.10 |= 2 x Intel Xeon Gold 6138 . 6.73 |= Radeon RX 580 ............ 174.90 |===================== Radeon R9 Fury ........... 330.70 |======================================== Radeon RX Vega 56 ........ 304.70 |===================================== Radeon RX Vega 64 ........ 363.90 |============================================ GeForce GTX 980 Ti ....... 216.50 |========================== GeForce GTX 1050 ......... 87.10 |=========== GeForce GTX 1050 Ti ...... 86.80 |========== GeForce GTX 1060 ......... 139.10 |================= GeForce GTX 1070 ......... 186.70 |======================= GeForce GTX 1070 Ti ...... 186.50 |======================= GeForce GTX 1080 ......... 208.90 |========================= GeForce GTX 1080 Ti ...... 317.10 |====================================== cl-mem 2017-01-13 Benchmark: Read GB/s > Higher Is Better AMD EPYC 7601 ............ 10.27 |= 2 x Intel Xeon Gold 6138 . 12.07 |= Radeon RX 580 ............ 208.40 |==================== Radeon R9 Fury ........... 448.20 |============================================ Radeon RX Vega 56 ........ 339.10 |================================= Radeon RX Vega 64 ........ 397.20 |======================================= GeForce GTX 980 Ti ....... 266.20 |========================== GeForce GTX 1050 ......... 95.10 |========= GeForce GTX 1050 Ti ...... 94.20 |========= GeForce GTX 1060 ......... 153.40 |=============== GeForce GTX 1070 ......... 205.40 |==================== GeForce GTX 1070 Ti ...... 205.50 |==================== GeForce GTX 1080 ......... 229.00 |====================== GeForce GTX 1080 Ti ...... 338.10 |================================= cl-mem 2017-01-13 Benchmark: Write GB/s > Higher Is Better AMD EPYC 7601 ............ 4.93 |= 2 x Intel Xeon Gold 6138 . 3.47 | Radeon RX 580 ............ 177.30 |==================== Radeon R9 Fury ........... 381.60 |============================================ Radeon RX Vega 56 ........ 321.90 |===================================== Radeon RX Vega 64 ........ 378.20 |============================================ GeForce GTX 980 Ti ....... 238.80 |============================ GeForce GTX 1050 ......... 86.30 |========== GeForce GTX 1050 Ti ...... 85.20 |========== GeForce GTX 1060 ......... 140.00 |================ GeForce GTX 1070 ......... 191.80 |====================== GeForce GTX 1070 Ti ...... 190.60 |====================== GeForce GTX 1080 ......... 215.60 |========================= GeForce GTX 1080 Ti ...... 336.30 |======================================= JuliaGPU 1.2pts1 OpenCL Samples/sec > Higher Is Better AMD EPYC 7601 ............ 28355131.50 |=== 2 x Intel Xeon Gold 6138 . 30799704.97 |=== Radeon RX 580 ............ 248225442.30 |=========================== Radeon R9 Fury ........... 269983077.70 |============================= Radeon RX Vega 56 ........ 331818340.60 |==================================== Radeon RX Vega 64 ........ 348619117.20 |====================================== GeForce GTX 980 Ti ....... 138987453.50 |=============== GeForce GTX 1050 ......... 67157808.00 |======= GeForce GTX 1050 Ti ...... 81114034.50 |========= GeForce GTX 1060 ......... 121054859.20 |============= GeForce GTX 1070 ......... 149885098.20 |================ GeForce GTX 1070 Ti ...... 171465266.00 |=================== GeForce GTX 1080 ......... 176402393.60 |=================== GeForce GTX 1080 Ti ...... 202544036.40 |======================