Hi everyone. I just started working with OpenCL a few weeks ago using pyopencl on a laptop at work with an Intel CPU and Nvidia GPU, and am now trying to get to work on my home machine with an AMD FX 8150 CPU and two HD6950 GPUs. I have installed version 3.0 of the APP SDK and the latest drivers available for the GPUs. Additional system details are below:
Windows 10 64-bit
Radeon Software Version: 16.2.1
Radeon Software Edition: Crimson
Radeon Settings Version: 2016.0226.1531.27895
Driver Packaging Version: 15.301.1901-160226a-318062E
OpenCL™ Version: 2.0.4.0
The main issue I am having is that when I try to run the samples, such as Mandelbrot.exe, I get a the following messages which indicate that the GPUs are not detected:
Platform 0: Intel(R) Corporation
Platform 1: Advanced Micro Devices, Inc.
GPU not found. Falling back to CPU device
Platform found: Advanced Micro Devices, Inc.
Selected Platform Vendor: Advanced Micro Devices, Inc.
Device 0: AMD FX(tm)-8150 Eight-Core Processor ....etc
The sample runs, but after completely uninstalling the APP SDK and display drivers, and then reinstalling the SDK first then display drivers, the GPUs are still not detected.
I suspect this is also why I cannot import pyopencl without getting the error below. After installing the Intel OpenCL SDK and drivers, I was able to import the package, but but the GPUs were still not listed as devices under the AMD platform. I have since uninstalled the Intel SDK and drivers, and reinstalled the APP SDK and display drivers. Is OpenCL just not supported on these GPUs anymore? Would an older version of the SDK work? I understand these GPUs are somewhat old and I am running beta drivers that are not supported anymore, so maybe there isn't a good solution. But I'd really appreciate any help getting this set up so I can get my feet wet learning OpenCL with the hardware I already have if possible.
Thank you!
pyopenCL error:
Python 3.6.3 |Anaconda custom (64-bit)| (default, Oct 15 2017, 03:27:45) [MSC v.1900 64 bit (AMD64)]
Type "copyright", "credits" or "license" for more information.
IPython 6.1.0 -- An enhanced Interactive Python.
Traceback (most recent call last):
File "<ipython-input-1-d80ce023813b>", line 1, in <module>
runfile('G:/pyopenclScripts/test1.py', wdir='G:/pyopenclScripts')
File "G:\Anaconda3\lib\site-packages\spyder\utils\site\sitecustomize.py", line 710, in runfile
execfile(filename, namespace)
File "G:\Anaconda3\lib\site-packages\spyder\utils\site\sitecustomize.py", line 101, in execfile
exec(compile(f.read(), filename, 'exec'), namespace)
File "G:/pyopenclScripts/test1.py", line 8, in <module>
import pyopencl as cl
File "G:\Anaconda3\lib\site-packages\pyopencl\__init__.py", line 37, in <module>
import pyopencl.cffi_cl as _cl
File "G:\Anaconda3\lib\site-packages\pyopencl\cffi_cl.py", line 39, in <module>
from pyopencl._cffi import ffi as _ffi
ImportError: DLL load failed: The specified procedure could not be found.
After installing the driver, please check the clinfo output. If the gpus not get detected there, then there is a problem with installation and any OCL sample will fail too.
Here is the driver for legacy gpu products like HD6950: Legacy
[I didn't use the beta version myself, however if I remember correctly, 15.7.1 worked fine when I installed it on a similar legacy card.]
Thank you. I checked clinfo and it also only shows the cpu. I will try the legacy 15.7.1 driver. Here is the clinfo output with my current configuration in case it has any other information that would be useful for debugging this:
Microsoft Windows [Version 10.0.15063]
(c) 2017 Microsoft Corporation. All rights reserved.
C:\Users\Brian>clinfo
Number of platforms: 1
Platform Profile: FULL_PROFILE
Platform Version: OpenCL 2.0 AMD-APP (1912.5)
Platform Name: AMD Accelerated Parallel Processing
Platform Vendor: Advanced Micro Devices, Inc.
Platform Extensions: cl_khr_icd cl_khr_d3d10_sharing cl_khr_d3d11_sharing cl_khr_dx9_media_sharing cl_amd_event_callback cl_amd_offline_devices
Platform Name: AMD Accelerated Parallel Processing
Number of devices: 1
Device Type: CL_DEVICE_TYPE_CPU
Vendor ID: 1002h
Board name:
Max compute units: 8
Max work items dimensions: 3
Max work items[0]: 1024
Max work items[1]: 1024
Max work items[2]: 1024
Max work group size: 1024
Preferred vector width char: 16
Preferred vector width short: 8
Preferred vector width int: 4
Preferred vector width long: 2
Preferred vector width float: 8
Preferred vector width double: 4
Native vector width char: 16
Native vector width short: 8
Native vector width int: 4
Native vector width long: 2
Native vector width float: 8
Native vector width double: 4
Max clock frequency: 4214Mhz
Address bits: 64
Max memory allocation: 4280635392
Image support: Yes
Max number of images read arguments: 128
Max number of images write arguments: 64
Max image 2D width: 8192
Max image 2D height: 8192
Max image 3D width: 2048
Max image 3D height: 2048
Max image 3D depth: 2048
Max samplers within kernel: 16
Max size of kernel argument: 4096
Alignment (bits) of base address: 1024
Minimum alignment (bytes) for any datatype: 128
Single precision floating point capability
Denorms: Yes
Quiet NaNs: Yes
Round to nearest even: Yes
Round to zero: Yes
Round to +ve and infinity: Yes
IEEE754-2008 fused multiply-add: Yes
Cache type: Read/Write
Cache line size: 64
Cache size: 16384
Global memory size: 17122541568
Constant buffer size: 65536
Max number of constant args: 8
Local memory type: Global
Local memory size: 32768
Max pipe arguments: 16
Max pipe active reservations: 16
Max pipe packet size: 4280635392
Max global variable size: 1879048192
Max global variable preferred total size: 1879048192
Max read/write image args: 64
Max on device events: 0
Queue on device max size: 0
Max on device queues: 0
Queue on device preferred size: 0
SVM capabilities:
Coarse grain buffer: No
Fine grain buffer: No
Fine grain system: No
Atomics: No
Preferred platform atomic alignment: 0
Preferred global atomic alignment: 0
Preferred local atomic alignment: 0
Kernel Preferred work group size multiple: 1
Error correction support: 0
Unified memory for Host and Device: 1
Profiling timer resolution: 69
Device endianess: Little
Available: Yes
Compiler available: Yes
Execution capabilities:
Execute OpenCL kernels: Yes
Execute native function: Yes
Queue on Host properties:
Out-of-Order: No
Profiling : Yes
Queue on Device properties:
Out-of-Order: No
Profiling : No
Platform ID: 00007FF958D0E188
Name: AMD FX(tm)-8150 Eight-Core Processor
Vendor: AuthenticAMD
Device OpenCL C version: OpenCL C 1.2
Driver version: 1912.5 (sse2,avx,fma4)
Profile: FULL_PROFILE
Version: OpenCL 1.2 AMD-APP (1912.5)
Extensions: cl_khr_fp64 cl_amd_fp64 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_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_gl_sharing cl_ext_device_fission cl_amd_device_attribute_query cl_amd_vec3 cl_amd_printf cl_amd_media_ops cl_amd_media_ops2 cl_amd_popcnt cl_khr_d3d10_sharing cl_khr_spir cl_khr_gl_event
C:\Users\Brian>
Thanks! I confirmed the GPUs are detected with the 15.7.1 legacy driver, but not with the 16.2.1 crimson beta driver. Maybe I can set up a separate partition with the older drivers to work with openCL and keep the newer drivers for gaming. I should get used to developing in a linux environment anyway, since that is most likely what I'd be working in in the future. This could be an opportunity to learn both.