AnsweredAssumed Answered

HD 6670 opencl GPU not recognised on windws 10

Question asked by diegoman on Oct 1, 2016
Latest reply on Oct 18, 2017 by mirh

My AMD HD 6670 should support opencl 1.1 according to the specifications, but both Caps Viewer and LuxMark show

only CPU  support for opencl, not GPU.

I have installed catalyst 15.7.1 and AMD APP SDK 3.0 on windows 10 64bit.


This is clinfo output:


Number of platforms: 1

  Platform Profile: FULL_PROFILE

  Platform Version:                      OpenCL 2.0 AMD-APP (1800.8)

  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:                     4

  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:                   3311Mhz

  Address bits:                          64

  Max memory allocation:                 2147483648

  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:                            32768

  Global memory size:                    8565694464

  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:                  2147483648

  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:            309

  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:                           00007FFA3446F180

  Name:                                           Intel(R) Core(TM) i5-2500 CPU @ 3.30GHz

  Vendor:                          GenuineIntel

  Device OpenCL C version:               OpenCL C 1.2

  Driver version:                  1800.8 (sse2,avx)

  Profile:                         FULL_PROFILE

  Version:                         OpenCL 1.2 AMD-APP (1800.8)

  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


What should I do to use GPU with openCL?

Can anyone please give me advice? Thanks.