OpenCL Radeon Open-Source vs. Catalyst, hd7970 Radeon R600g and RadeonSI Gallium3D Clover OpenCL benchmarks with clpeak using Linux 4.5 + Mesa 11.2-devel. Compared to Catalyst. Benchmark results by Michael Larabel of Phoronix.com for a future article. hd7970: Intel Core i5-3330 testing with a ASRock H61 Pro BTC and Intel HD 2500 3072MB on SUSE LINUX 42.1 via the Phoronix Test Suite. Open-Source: HD 6870: Processor: Intel Core i7-5960X @ 3.50GHz (16 Cores), Motherboard: Gigabyte X99-UD4-CF, Chipset: Intel Xeon E7 v3/Xeon, Memory: 16384MB, Disk: 2 x 120GB SanDisk SDSSDA12 + 120GB Samsung SSD 850, Graphics: Sapphire AMD Radeon HD 6870 1024MB, Audio: Realtek ALC1150, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 15.10, Kernel: 4.5.0-rc2amdgpu-pp+ (x86_64) 20160201, Desktop: Unity, Display Server: X Server 1.17.2, Display Driver: radeon 7.6.99, OpenGL: 3.3 Mesa 11.2.0-devel (padoka PPA) Gallium 0.4, Compiler: GCC 5.2.1 20151010, File-System: ext4, Screen Resolution: 3840x2160 Open-Source: HD 7950: Processor: Intel Core i7-5960X @ 3.50GHz (16 Cores), Motherboard: Gigabyte X99-UD4-CF, Chipset: Intel Xeon E7 v3/Xeon, Memory: 16384MB, Disk: 2 x 120GB SanDisk SDSSDA12 + 120GB Samsung SSD 850, Graphics: XFX AMD Radeon HD 7950/8950 OEM / R9 280 3072MB, Audio: Realtek ALC1150, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 15.10, Kernel: 4.5.0-rc2amdgpu-pp+ (x86_64) 20160201, Desktop: Unity, Display Server: X Server 1.17.2, Display Driver: radeon 7.6.99, OpenGL: 4.1 Mesa 11.2.0-devel (padoka PPA) Gallium 0.4, Compiler: GCC 5.2.1 20151010, File-System: ext4, Screen Resolution: 3840x2160 Open-Source: R9 285: Processor: Intel Core i7-5960X @ 3.50GHz (16 Cores), Motherboard: Gigabyte X99-UD4-CF, Chipset: Intel Xeon E7 v3/Xeon, Memory: 16384MB, Disk: 2 x 120GB SanDisk SDSSDA12 + 120GB Samsung SSD 850, Graphics: XFX AMD Radeon R9 285/380 2048MB, Audio: Realtek ALC1150, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 15.10, Kernel: 4.5.0-rc2amdgpu-pp+ (x86_64) 20160201, Desktop: Unity, Display Server: X Server 1.17.2, Display Driver: amdgpu 1.0.99, OpenGL: 4.1 Mesa 11.2.0-devel (padoka PPA) Gallium 0.4, Compiler: GCC 5.2.1 20151010, File-System: ext4, Screen Resolution: 3840x2160 Open-Source: R9 290: Processor: Intel Core i7-5960X @ 3.50GHz (16 Cores), Motherboard: Gigabyte X99-UD4-CF, Chipset: Intel Xeon E7 v3/Xeon, Memory: 16384MB, Disk: 2 x 120GB SanDisk SDSSDA12 + 120GB Samsung SSD 850, Graphics: XFX AMD Radeon R9 290 4096MB, Audio: Realtek ALC1150, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 15.10, Kernel: 4.5.0-rc2amdgpu-pp+ (x86_64) 20160201, Desktop: Unity, Display Server: X Server 1.17.2, Display Driver: radeon 7.6.99, OpenGL: 4.1 Mesa 11.2.0-devel (padoka PPA) Gallium 0.4, Compiler: GCC 5.2.1 20151010, File-System: ext4, Screen Resolution: 3840x2160 Open-Source: R7 370: Processor: Intel Core i7-5960X @ 3.50GHz (16 Cores), Motherboard: Gigabyte X99-UD4-CF, Chipset: Intel Xeon E7 v3/Xeon, Memory: 16384MB, Disk: 2 x 120GB SanDisk SDSSDA12 + 120GB Samsung SSD 850, Graphics: MSI AMD Radeon R7 370 / R9 270/370 OEM 4096MB, Audio: Realtek ALC1150, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 15.10, Kernel: 4.5.0-rc2amdgpu-pp+ (x86_64) 20160201, Desktop: Unity, Display Server: X Server 1.17.2, Display Driver: radeon 7.6.99, OpenGL: 4.1 Mesa 11.2.0-devel (padoka PPA) Gallium 0.4, Compiler: GCC 5.2.1 20151010, File-System: ext4, Screen Resolution: 3840x2160 Catalyst: HD 6870: Processor: Intel Core i7-5960X @ 3.50GHz (16 Cores), Motherboard: Gigabyte X99-UD4-CF, Chipset: Intel Xeon E7 v3/Xeon, Memory: 16384MB, Disk: 2 x 120GB SanDisk SDSSDA12 + 120GB Samsung SSD 850, Graphics: Sapphire AMD Radeon HD 6800 1024MB, Audio: Realtek ALC1150, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 15.10, Kernel: 4.2.0-27-generic (x86_64), Desktop: Unity, Display Server: X Server 1.17.2, Display Driver: fglrx 15.20.3, OpenGL: 4.5.13399, Compiler: GCC 5.2.1 20151010, File-System: ext4, Screen Resolution: 3840x2160 Catalyst: HD 7950: Processor: Intel Core i7-5960X @ 3.50GHz (16 Cores), Motherboard: Gigabyte X99-UD4-CF, Chipset: Intel Xeon E7 v3/Xeon, Memory: 16384MB, Disk: 2 x 120GB SanDisk SDSSDA12 + 120GB Samsung SSD 850, Graphics: XFX AMD Radeon HD 7900 3072MB, Audio: Realtek ALC1150, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 15.10, Kernel: 4.2.0-27-generic (x86_64), Desktop: Unity, Display Server: X Server 1.17.2, Display Driver: fglrx 15.20.3, OpenGL: 4.5.13399, Compiler: GCC 5.2.1 20151010, File-System: ext4, Screen Resolution: 3840x2160 Catalyst: R9 285: Processor: Intel Core i7-5960X @ 3.50GHz (16 Cores), Motherboard: Gigabyte X99-UD4-CF, Chipset: Intel Xeon E7 v3/Xeon, Memory: 16384MB, Disk: 2 x 120GB SanDisk SDSSDA12 + 120GB Samsung SSD 850, Graphics: Supported device 6939 2048MB, Audio: Realtek ALC1150, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 15.10, Kernel: 4.2.0-27-generic (x86_64), Desktop: Unity, Display Server: X Server 1.17.2, Display Driver: fglrx 15.20.3, OpenGL: 4.5.13399, Compiler: GCC 5.2.1 20151010, File-System: ext4, Screen Resolution: 3840x2160 Catalyst: R9 290: Processor: Intel Core i7-5960X @ 3.50GHz (16 Cores), Motherboard: Gigabyte X99-UD4-CF, Chipset: Intel Xeon E7 v3/Xeon, Memory: 16384MB, Disk: 2 x 120GB SanDisk SDSSDA12 + 120GB Samsung SSD 850, Graphics: Supported device 67B1 4096MB, Audio: Realtek ALC1150, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 15.10, Kernel: 4.2.0-27-generic (x86_64), Desktop: Unity, Display Server: X Server 1.17.2, Display Driver: fglrx 15.20.3, OpenGL: 4.5.13399, Compiler: GCC 5.2.1 20151010, File-System: ext4, Screen Resolution: 3840x2160 Catalyst: R7 370: Processor: Intel Core i7-5960X @ 3.50GHz (16 Cores), Motherboard: Gigabyte X99-UD4-CF, Chipset: Intel Xeon E7 v3/Xeon, Memory: 16384MB, Disk: 2 x 120GB SanDisk SDSSDA12 + 120GB Samsung SSD 850, Graphics: MSI AMD Radeon R7 370 4096MB, Audio: Realtek ALC1150, Monitor: DELL P2415Q, Network: Intel Connection OS: Ubuntu 15.10, Kernel: 4.2.0-27-generic (x86_64), Desktop: Unity, Display Server: X Server 1.17.2, Display Driver: fglrx 15.20.3, OpenGL: 4.5.13399, Compiler: GCC 5.2.1 20151010, File-System: ext4, Screen Resolution: 3840x2160 hd7970: Processor: Intel Core i5-3330 @ 3.20GHz (4 Cores), Motherboard: ASRock H61 Pro BTC, Chipset: Intel Xeon E3-1200 v2/3rd, Memory: 8192MB, Disk: 65GB, Graphics: Intel HD 2500 3072MB (1050MHz), Audio: VIA VT1705, Monitor: H225W, Network: Realtek RTL8111/8168/8411 OS: SUSE LINUX 42.1, Kernel: 4.5.0-rc1-3-default (x86_64), Desktop: KDE Frameworks 5, Display Server: X Server 1.18.99.1, Display Driver: intel 2.99.917, OpenGL: 4.1 Mesa 11.2.0-devel (git-43a401a pontostroy:X11) Gallium 0.4, Compiler: GCC 4.8.5 + Clang 3.9.0 + LLVM 3.9.0svn, File-System: xfs, Screen Resolution: 1920x1080 clpeak OpenCL Test: Kernel Latency us < Lower Is Better Open-Source: HD 6870 . 145.85 |================================================ Open-Source: HD 7950 . 105.54 |=================================== Open-Source: R9 285 .. 143.39 |=============================================== Open-Source: R9 290 .. 118.71 |======================================= Open-Source: R7 370 .. 110.24 |==================================== Catalyst: HD 6870 .... 24.90 |======== Catalyst: HD 7950 .... 34.03 |=========== Catalyst: R9 285 ..... 26.78 |========= Catalyst: R9 290 ..... 37.77 |============ Catalyst: R7 370 ..... 35.42 |============ hd7970 ............... 69.45 |======================= clpeak OpenCL Test: Integer Compute INT GIOPS > Higher Is Better Open-Source: HD 6870 . 182.44 |===== Open-Source: HD 7950 . 491.62 |=============== Open-Source: R9 285 .. 650.80 |=================== Open-Source: R9 290 .. 1578.01 |=============================================== Open-Source: R7 370 .. 418.74 |============ Catalyst: HD 6870 .... 199.38 |====== Catalyst: HD 7950 .... 516.74 |=============== Catalyst: R9 285 ..... 647.08 |=================== Catalyst: R9 290 ..... 1591.74 |=============================================== Catalyst: R7 370 ..... 415.67 |============ hd7970 ............... 847.70 |========================= clpeak OpenCL Test: Single-Precision Float GFLOPS > Higher Is Better Open-Source: HD 6870 . 360.92 |==== Open-Source: HD 7950 . 2345.12 |======================= Open-Source: R9 285 .. 3102.60 |=============================== Open-Source: R9 290 .. 4608.55 |============================================== Open-Source: R7 370 .. 2031.91 |==================== Catalyst: HD 6870 .... 398.29 |==== Catalyst: HD 7950 .... 2292.54 |======================= Catalyst: R9 285 ..... 3177.62 |================================ Catalyst: R9 290 ..... 4721.18 |=============================================== Catalyst: R7 370 ..... 2095.91 |===================== hd7970 ............... 4149.22 |========================================= clpeak OpenCL Test: Double-Precision Double GFLOPS > Higher Is Better Open-Source: HD 7950 . 547.03 |======================== Open-Source: R9 285 .. 204.75 |========= Open-Source: R9 290 .. 596.55 |========================== Open-Source: R7 370 .. 131.49 |====== Catalyst: HD 7950 .... 536.84 |======================== Catalyst: R9 285 ..... 202.71 |========= Catalyst: R9 290 ..... 598.81 |=========================== Catalyst: R7 370 ..... 128.73 |====== hd7970 ............... 1058.38 |=============================================== clpeak OpenCL Test: Transfer Bandwidth enqueueReadBuffer GBPS > Higher Is Better Open-Source: HD 6870 . 3.58 |================ Open-Source: HD 7950 . 10.40 |================================================ Open-Source: R9 285 .. 10.09 |============================================== Open-Source: R9 290 .. 9.82 |============================================= Open-Source: R7 370 .. 10.41 |================================================ Catalyst: HD 6870 .... 8.80 |======================================== Catalyst: HD 7950 .... 10.47 |================================================ Catalyst: R9 285 ..... 9.00 |========================================= Catalyst: R9 290 ..... 10.64 |================================================= Catalyst: R7 370 ..... 10.67 |================================================= hd7970 ............... 5.83 |=========================== clpeak OpenCL Test: Transfer Bandwidth enqueueWriteBuffer GBPS > Higher Is Better Open-Source: HD 6870 . 5.26 |================ Open-Source: HD 7950 . 15.72 |================================================= Open-Source: R9 285 .. 14.82 |============================================== Open-Source: R9 290 .. 12.32 |====================================== Open-Source: R7 370 .. 15.77 |================================================= Catalyst: HD 6870 .... 9.49 |============================= Catalyst: HD 7950 .... 13.88 |=========================================== Catalyst: R9 285 ..... 13.21 |========================================= Catalyst: R9 290 ..... 14.81 |============================================== Catalyst: R7 370 ..... 14.97 |=============================================== hd7970 ............... 7.99 |========================= clpeak OpenCL Test: Global Memory Bandwidth GBPS > Higher Is Better hd7970 . 168.85 |==============================================================