# Finally OpenCL 1.1 driver was released! **URL:** \ **Category:** CUDA Programming and Performance\ **Created:** [August 3, 2011, 4:59am UTC](https://forums.developer.nvidia.com/t/finally-opencl-1-1-driver-was-released/23786 "2011-08-03T04:59:12Z")\ **Posts on this page:** 11\ **Page:** 1
**Author:** ![ZHAO\_Peng](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@ZHAO\_Peng](https://forums.developer.nvidia.com/u/ZHAO_Peng)\ **Post date:** [August 3, 2011, 4:59am UTC](https://forums.developer.nvidia.com/t/finally-opencl-1-1-driver-was-released/23786/1 "2011-08-03T04:59:12Z")
After waiting more than one year, the official OpenCL 1.1 driver(280.13 for linux, 280.19 beta for windows) was released. Many times I think Nvidia has given up the support of OpenCL. Now we should set our minds at rest. [External Image](http://forums.nvidia.com/public/style_emoticons/<#EMO_DIR#>/smile.gif "Image hosted on another site. Click to open in a new tab.") This is the result of sample oclDeviceQuery on my desktop: ./oclDeviceQuery Starting… OpenCL SW Info: CL\_PLATFORM\_NAME: NVIDIA CUDA CL\_PLATFORM\_VERSION: OpenCL 1.1 CUDA 4.0.1 OpenCL SDK Revision: 7027912 OpenCL Device Info: 1 devices found supporting OpenCL: * * * ## Device GeForce GTX 570 CL\_DEVICE\_NAME: GeForce GTX 570 CL\_DEVICE\_VENDOR: NVIDIA Corporation CL\_DRIVER\_VERSION: 280.13 CL\_DEVICE\_VERSION: OpenCL 1.1 CUDA CL\_DEVICE\_OPENCL\_C\_VERSION: OpenCL C 1.1 CL\_DEVICE\_TYPE: CL\_DEVICE\_TYPE\_GPU CL\_DEVICE\_MAX\_COMPUTE\_UNITS: 15 CL\_DEVICE\_MAX\_WORK\_ITEM\_DIMENSIONS: 3 CL\_DEVICE\_MAX\_WORK\_ITEM\_SIZES: 1024 / 1024 / 64 CL\_DEVICE\_MAX\_WORK\_GROUP\_SIZE: 1024 CL\_DEVICE\_MAX\_CLOCK\_FREQUENCY: 1464 MHz CL\_DEVICE\_ADDRESS\_BITS: 32 CL\_DEVICE\_MAX\_MEM\_ALLOC\_SIZE: 319 MByte CL\_DEVICE\_GLOBAL\_MEM\_SIZE: 1279 MByte CL\_DEVICE\_ERROR\_CORRECTION\_SUPPORT: no CL\_DEVICE\_LOCAL\_MEM\_TYPE: local CL\_DEVICE\_LOCAL\_MEM\_SIZE: 48 KByte CL\_DEVICE\_MAX\_CONSTANT\_BUFFER\_SIZE: 64 KByte CL\_DEVICE\_QUEUE\_PROPERTIES: CL\_QUEUE\_OUT\_OF\_ORDER\_EXEC\_MODE\_ENABLE CL\_DEVICE\_QUEUE\_PROPERTIES: CL\_QUEUE\_PROFILING\_ENABLE CL\_DEVICE\_IMAGE\_SUPPORT: 1 CL\_DEVICE\_MAX\_READ\_IMAGE\_ARGS: 128 CL\_DEVICE\_MAX\_WRITE\_IMAGE\_ARGS: 8 CL\_DEVICE\_SINGLE\_FP\_CONFIG: denorms INF-quietNaNs round-to-nearest round-to-zero round-to-inf fma CL\_DEVICE\_IMAGE 2D\_MAX\_WIDTH 16384 2D\_MAX\_HEIGHT 16384 3D\_MAX\_WIDTH 2048 3D\_MAX\_HEIGHT 2048 3D\_MAX\_DEPTH 2048 CL\_DEVICE\_EXTENSIONS: cl\_khr\_byte\_addressable\_store cl\_khr\_icd cl\_khr\_gl\_sharing cl\_nv\_compiler\_options cl\_nv\_device\_attribute\_query cl\_nv\_pragma\_unroll cl\_khr\_global\_int32\_base\_atomics cl\_khr\_global\_int32\_extended\_atomics cl\_khr\_local\_int32\_base\_atomics cl\_khr\_local\_int32\_extended\_atomics cl\_khr\_fp64 CL\_DEVICE\_COMPUTE\_CAPABILITY\_NV: 2.0 NUMBER OF MULTIPROCESSORS: 15 NUMBER OF CUDA CORES: 480 CL\_DEVICE\_REGISTERS\_PER\_BLOCK\_NV: 32768 CL\_DEVICE\_WARP\_SIZE\_NV: 32 CL\_DEVICE\_GPU\_OVERLAP\_NV: CL\_TRUE CL\_DEVICE\_KERNEL\_EXEC\_TIMEOUT\_NV: CL\_TRUE CL\_DEVICE\_INTEGRATED\_MEMORY\_NV: CL\_FALSE CL\_DEVICE\_PREFERRED\_VECTOR\_WIDTH\_ CHAR 1, SHORT 1, INT 1, LONG 1, FLOAT 1, DOUBLE 1 * * * ## 2D Image Formats Supported (71) # Channel Order Channel Type 1 CL\_R CL\_FLOAT 2 CL\_R CL\_HALF\_FLOAT 3 CL\_R CL\_UNORM\_INT8 4 CL\_R CL\_UNORM\_INT16 5 CL\_R CL\_SNORM\_INT16 6 CL\_R CL\_SIGNED\_INT8 7 CL\_R CL\_SIGNED\_INT16 8 CL\_R CL\_SIGNED\_INT32 9 CL\_R CL\_UNSIGNED\_INT8 10 CL\_R CL\_UNSIGNED\_INT16 11 CL\_R CL\_UNSIGNED\_INT32 12 CL\_A CL\_FLOAT 13 CL\_A CL\_HALF\_FLOAT 14 CL\_A CL\_UNORM\_INT8 15 CL\_A CL\_UNORM\_INT16 16 CL\_A CL\_SNORM\_INT16 17 CL\_A CL\_SIGNED\_INT8 18 CL\_A CL\_SIGNED\_INT16 19 CL\_A CL\_SIGNED\_INT32 20 CL\_A CL\_UNSIGNED\_INT8 21 CL\_A CL\_UNSIGNED\_INT16 22 CL\_A CL\_UNSIGNED\_INT32 23 CL\_RG CL\_FLOAT 24 CL\_RG CL\_HALF\_FLOAT 25 CL\_RG CL\_UNORM\_INT8 26 CL\_RG CL\_UNORM\_INT16 27 CL\_RG CL\_SNORM\_INT16 28 CL\_RG CL\_SIGNED\_INT8 29 CL\_RG CL\_SIGNED\_INT16 30 CL\_RG CL\_SIGNED\_INT32 31 CL\_RG CL\_UNSIGNED\_INT8 32 CL\_RG CL\_UNSIGNED\_INT16 33 CL\_RG CL\_UNSIGNED\_INT32 34 CL\_RA CL\_FLOAT 35 CL\_RA CL\_HALF\_FLOAT 36 CL\_RA CL\_UNORM\_INT8 37 CL\_RA CL\_UNORM\_INT16 38 CL\_RA CL\_SNORM\_INT16 39 CL\_RA CL\_SIGNED\_INT8 40 CL\_RA CL\_SIGNED\_INT16 41 CL\_RA CL\_SIGNED\_INT32 42 CL\_RA CL\_UNSIGNED\_INT8 43 CL\_RA CL\_UNSIGNED\_INT16 44 CL\_RA CL\_UNSIGNED\_INT32 45 CL\_RGBA CL\_FLOAT 46 CL\_RGBA CL\_HALF\_FLOAT 47 CL\_RGBA CL\_UNORM\_INT8 48 CL\_RGBA CL\_UNORM\_INT16 49 CL\_RGBA CL\_SNORM\_INT16 50 CL\_RGBA CL\_SIGNED\_INT8 51 CL\_RGBA CL\_SIGNED\_INT16 52 CL\_RGBA CL\_SIGNED\_INT32 53 CL\_RGBA CL\_UNSIGNED\_INT8 54 CL\_RGBA CL\_UNSIGNED\_INT16 55 CL\_RGBA CL\_UNSIGNED\_INT32 56 CL\_BGRA CL\_UNORM\_INT8 57 CL\_BGRA CL\_SIGNED\_INT8 58 CL\_BGRA CL\_UNSIGNED\_INT8 59 CL\_ARGB CL\_UNORM\_INT8 60 CL\_ARGB CL\_SIGNED\_INT8 61 CL\_ARGB CL\_UNSIGNED\_INT8 62 CL\_INTENSITY CL\_FLOAT 63 CL\_INTENSITY CL\_HALF\_FLOAT 64 CL\_INTENSITY CL\_UNORM\_INT8 65 CL\_INTENSITY CL\_UNORM\_INT16 66 CL\_INTENSITY CL\_SNORM\_INT16 67 CL\_LUMINANCE CL\_FLOAT 68 CL\_LUMINANCE CL\_HALF\_FLOAT 69 CL\_LUMINANCE CL\_UNORM\_INT8 70 CL\_LUMINANCE CL\_UNORM\_INT16 71 CL\_LUMINANCE CL\_SNORM\_INT16 * * * ## 3D Image Formats Supported (71) # Channel Order Channel Type 1 CL\_R CL\_FLOAT 2 CL\_R CL\_HALF\_FLOAT 3 CL\_R CL\_UNORM\_INT8 4 CL\_R CL\_UNORM\_INT16 5 CL\_R CL\_SNORM\_INT16 6 CL\_R CL\_SIGNED\_INT8 7 CL\_R CL\_SIGNED\_INT16 8 CL\_R CL\_SIGNED\_INT32 9 CL\_R CL\_UNSIGNED\_INT8 10 CL\_R CL\_UNSIGNED\_INT16 11 CL\_R CL\_UNSIGNED\_INT32 12 CL\_A CL\_FLOAT 13 CL\_A CL\_HALF\_FLOAT 14 CL\_A CL\_UNORM\_INT8 15 CL\_A CL\_UNORM\_INT16 16 CL\_A CL\_SNORM\_INT16 17 CL\_A CL\_SIGNED\_INT8 18 CL\_A CL\_SIGNED\_INT16 19 CL\_A CL\_SIGNED\_INT32 20 CL\_A CL\_UNSIGNED\_INT8 21 CL\_A CL\_UNSIGNED\_INT16 22 CL\_A CL\_UNSIGNED\_INT32 23 CL\_RG CL\_FLOAT 24 CL\_RG CL\_HALF\_FLOAT 25 CL\_RG CL\_UNORM\_INT8 26 CL\_RG CL\_UNORM\_INT16 27 CL\_RG CL\_SNORM\_INT16 28 CL\_RG CL\_SIGNED\_INT8 29 CL\_RG CL\_SIGNED\_INT16 30 CL\_RG CL\_SIGNED\_INT32 31 CL\_RG CL\_UNSIGNED\_INT8 32 CL\_RG CL\_UNSIGNED\_INT16 33 CL\_RG CL\_UNSIGNED\_INT32 34 CL\_RA CL\_FLOAT 35 CL\_RA CL\_HALF\_FLOAT 36 CL\_RA CL\_UNORM\_INT8 37 CL\_RA CL\_UNORM\_INT16 38 CL\_RA CL\_SNORM\_INT16 39 CL\_RA CL\_SIGNED\_INT8 40 CL\_RA CL\_SIGNED\_INT16 41 CL\_RA CL\_SIGNED\_INT32 42 CL\_RA CL\_UNSIGNED\_INT8 43 CL\_RA CL\_UNSIGNED\_INT16 44 CL\_RA CL\_UNSIGNED\_INT32 45 CL\_RGBA CL\_FLOAT 46 CL\_RGBA CL\_HALF\_FLOAT 47 CL\_RGBA CL\_UNORM\_INT8 48 CL\_RGBA CL\_UNORM\_INT16 49 CL\_RGBA CL\_SNORM\_INT16 50 CL\_RGBA CL\_SIGNED\_INT8 51 CL\_RGBA CL\_SIGNED\_INT16 52 CL\_RGBA CL\_SIGNED\_INT32 53 CL\_RGBA CL\_UNSIGNED\_INT8 54 CL\_RGBA CL\_UNSIGNED\_INT16 55 CL\_RGBA CL\_UNSIGNED\_INT32 56 CL\_BGRA CL\_UNORM\_INT8 57 CL\_BGRA CL\_SIGNED\_INT8 58 CL\_BGRA CL\_UNSIGNED\_INT8 59 CL\_ARGB CL\_UNORM\_INT8 60 CL\_ARGB CL\_SIGNED\_INT8 61 CL\_ARGB CL\_UNSIGNED\_INT8 62 CL\_INTENSITY CL\_FLOAT 63 CL\_INTENSITY CL\_HALF\_FLOAT 64 CL\_INTENSITY CL\_UNORM\_INT8 65 CL\_INTENSITY CL\_UNORM\_INT16 66 CL\_INTENSITY CL\_SNORM\_INT16 67 CL\_LUMINANCE CL\_FLOAT 68 CL\_LUMINANCE CL\_HALF\_FLOAT 69 CL\_LUMINANCE CL\_UNORM\_INT8 70 CL\_LUMINANCE CL\_UNORM\_INT16 71 CL\_LUMINANCE CL\_SNORM\_INT16 oclDeviceQuery, Platform Name = NVIDIA CUDA, Platform Version = OpenCL 1.1 CUDA 4.0.1, SDK Revision = 7027912, NumDevs = 1, Device = GeForce GTX 570 System Info: Local Time/Date = 12:38:09, 08/03/2011 CPU Name: Intel(R) Core™ i5-2500K CPU @ 3.30GHz # of CPU processors: 4 Linux version 2.6.37.6-0.7-desktop (geeko@buildhost) (gcc version 4.5.1 20101208 [gcc-4\_5-branch revision 167585] (SUSE Linux) ) #1 SMP PREEMPT 2011-07-21 02:17:24 +0200 ---
**Author:** ![eyebex](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@eyebex](https://forums.developer.nvidia.com/u/eyebex)\ **Post date:** [August 3, 2011, 6:38am UTC](https://forums.developer.nvidia.com/t/finally-opencl-1-1-driver-was-released/23786/2 "2011-08-03T06:38:44Z")
And still, CL\_DEVICE\_MAX\_MEM\_ALLOC\_SIZE is just 1/4 of CL\_DEVICE\_GLOBAL\_MEM\_SIZE :-/ ---
**Author:** ![Maxim\_Milakov](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/maxim_milakov/32/15909_2.png) [@Maxim\_Milakov](https://forums.developer.nvidia.com/u/Maxim_Milakov)\ **Post date:** [August 3, 2011, 8:09am UTC](https://forums.developer.nvidia.com/t/finally-opencl-1-1-driver-was-released/23786/3 "2011-08-03T08:09:34Z")
> [@](#): > > After waiting more than one year, the official OpenCL 1.1 driver(280.13 for linux, 280.19 beta for windows) was released. Many times I think Nvidia has given up the support of OpenCL. Now we should set our minds at rest. [External Image](http://forums.nvidia.com/public/style_emoticons/<#EMO_DIR#>/smile.gif "Image hosted on another site. Click to open in a new tab.") Well, it is great news indeed :) But the windows driver is still beta, and Developer Drivers for WinVista and Win7 is still 270.81. ---
**Author:** ![Martin\_Nilsson](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@Martin\_Nilsson](https://forums.developer.nvidia.com/u/Martin_Nilsson)\ **Post date:** [August 4, 2011, 2:28pm UTC](https://forums.developer.nvidia.com/t/finally-opencl-1-1-driver-was-released/23786/4 "2011-08-04T14:28:37Z")
I hear that there are some significant performance degradation with this driver, at least on Windows, [link.](http://www.gpu-tech.org/content.php/162-Nvidia-supports-OpenCL-1.1-with-Geforce-280.19-Beta-performance-suffers) Can anyone confirm? Is it the same on Linux? ---
**Author:** ![RyFo18](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@RyFo18](https://forums.developer.nvidia.com/u/RyFo18)\ **Post date:** [August 4, 2011, 4:55pm UTC](https://forums.developer.nvidia.com/t/finally-opencl-1-1-driver-was-released/23786/5 "2011-08-04T16:55:11Z")
> [@](#): > > I hear that there are some significant performance degradation with this driver, at least on Windows, link. Can anyone confirm? Is it the same on Linux? Haven’t noticed a degradation on Linux…Have noticed some speedup actually in the case of Multiple GPUs running simultaneously. Now watch what will happen, Khronos will release OpenCL 1.2 tomorrow [External Image](http://forums.nvidia.com/public/style_emoticons/<#EMO_DIR#>/biggrin.gif "Image hosted on another site. Click to open in a new tab.") ---
**Author:** ![RyFo18](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@RyFo18](https://forums.developer.nvidia.com/u/RyFo18)\ **Post date:** [August 4, 2011, 5:36pm UTC](https://forums.developer.nvidia.com/t/finally-opencl-1-1-driver-was-released/23786/6 "2011-08-04T17:36:58Z")
I take that back, I have noticed some slow downs…but have also noticed some speedups in the case of Multiple GPUs running simultaneously (clFinish is much faster in that case). ---
**Author:** ![bdelgado](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@bdelgado](https://forums.developer.nvidia.com/u/bdelgado)\ **Post date:** [August 4, 2011, 6:07pm UTC](https://forums.developer.nvidia.com/t/finally-opencl-1-1-driver-was-released/23786/7 "2011-08-04T18:07:47Z")
Pardon my ignorance of OpenCL but what are the hardware requirements for OpenCL 1.1 support? E.g. I have a Tesla C1060, does that support OpenCL 1.1? ---
**Author:** ![ZHAO\_Peng](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@ZHAO\_Peng](https://forums.developer.nvidia.com/u/ZHAO_Peng)\ **Post date:** [August 5, 2011, 8:55am UTC](https://forums.developer.nvidia.com/t/finally-opencl-1-1-driver-was-released/23786/8 "2011-08-05T08:55:54Z")
> [@](#): > > I hear that there are some significant performance degradation with this driver, at least on Windows, link. Can anyone confirm? Is it the same on Linux? I have not run any benchmark software of OpenCL on Linux. But for my little median filter sample the performance is improved slightly. After so long development, maybe the implementation of Nvidia has been changed greatly. And they should update the performance guide sometime. After all, we should be glad to see that Nvidia have not killed OpenCL. [External Image](http://forums.nvidia.com/public/style_emoticons/<#EMO_DIR#>/biggrin.gif "Image hosted on another site. Click to open in a new tab.") ---
**Author:** ![ZHAO\_Peng](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@ZHAO\_Peng](https://forums.developer.nvidia.com/u/ZHAO_Peng)\ **Post date:** [August 5, 2011, 9:00am UTC](https://forums.developer.nvidia.com/t/finally-opencl-1-1-driver-was-released/23786/9 "2011-08-05T09:00:43Z")
> [@](#): > > Pardon my ignorance of OpenCL but what are the hardware requirements for OpenCL 1.1 support? E.g. I have a Tesla C1060, does that support OpenCL 1.1? Yes, Tesla C1060 supports OpenCL1.1 with the Quadro/Tesla Driver v280.19 WHQL. If I make no mistake, any hardware which supports CUDA should support OpenCL 1.1 with the latest driver. ---
**Author:** ![NCC-1701D](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@NCC-1701D](https://forums.developer.nvidia.com/u/NCC-1701D)\ **Post date:** [August 5, 2011, 9:05am UTC](https://forums.developer.nvidia.com/t/finally-opencl-1-1-driver-was-released/23786/10 "2011-08-05T09:05:12Z")
> [@](#): > > I hear that there are some significant performance degradation with this driver, at least on Windows, link. Can anyone confirm? Is it the same on Linux? I am seeing upto 3x slowdown in linux (with 280.13). Mainly with compute heavy kernels, especially the one with a lot of SFU ops. ---
**Author:** ![eyebex](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@eyebex](https://forums.developer.nvidia.com/u/eyebex)\ **Post date:** [August 19, 2011, 7:12am UTC](https://forums.developer.nvidia.com/t/finally-opencl-1-1-driver-was-released/23786/11 "2011-08-19T07:12:13Z")
> [@](#): > > Yes, Tesla C1060 supports OpenCL1.1 with the Quadro/Tesla Driver v280.19 WHQL. If I make no mistake, any hardware which supports CUDA should support OpenCL 1.1 with the latest driver. I don’t think so. According to [this table](http://en.wikipedia.org/wiki/Nvidia_Tesla#Specifications_and_configurations) the C1060 has Compute Capability 1.3, which is the same as e.g. the GTX 285, which in turn is not an OpenCL 1.1 device (see [this post](http://forums.nvidia.com/index.php?showtopic=207323&view=findpost&p=1278308)).