Intel GPU support

[psychocrypt]

Add/Check OpenCl support for Intel GPUs.

#17 added support for GPUs without AMD media extensions, this means that Intel GPUs should be usable.

[psychocrypt]

Test Intel GPU on a i7-7500U

• system: Mint Linux 18.1
• download OpenCL™ 2.0 GPU/CPU driver package for Linux* (64-bit)

# convert *.rpm to *.dep
sudo apt-get install -y rpm alien libnuma1
fakeroot alien --to-deb intel-opencl-r4.1-61547.x86_64.rpm
fakeroot alien --to-deb intel-opencl-devel-r4.1-61547.x86_64.rpm
fakeroot alien --to-deb intel-opencl-cpu-r4.1-61547.x86_64.rpm  
# install dep packages
# compile xmr-stak-amd 

Everything is running without any errors (with OpenCL2):

2017-06-26 20:44:37] : New block detected.
[2017-06-26 20:44:46] : Result accepted by the pool.
[2017-06-26 20:44:56] : Result accepted by the pool.
[2017-06-26 20:45:27] : New block detected.
[2017-06-26 20:45:54] : Result accepted by the pool.
[2017-06-26 20:46:07] : New block detected.
HASHRATE REPORT
| ID |   10s |   60s |   15m |
|  0 | 84.8 | 85.5 | 85.8 |
---------------------------
Totals:   84.8 85.5 85.8 H/s
Highest:  86.5 H/s
RESULT REPORT
Difficulty       : 18000
Good results     : 9 / 9 (100.0 %)
Avg result time  : 175.3 sec
Pool-side hashes : 162000

Top 10 best results found:
|  0 |           344331 |  1 |           163503 |
|  2 |            86761 |  3 |            55549 |
|  4 |            51698 |  5 |            45328 |
|  6 |            34849 |  7 |            23566 |
|  8 |            19784 |  9 |                0 |

Error details:
Yay! No errors.

[fireice-uk]

Hold on, does that mean almost all Intel desktop CPUs can mine on their built-in GPU? Sweet =). Can you verify how that affects CPU mining rate? I would expect some shared circuits between the two.

[psychocrypt]

Yes it is possible but I have tested if a cpu miner and gpu miner can run in parallel. The result was that the cpu miner slows down from 130 to 30h/s.
The GPU of my laptop is running with 80 h/s (allown).
But all in all I need to run more tests to check why the cpu miner slows down if a gpu miner is running.

[arpu]

please show clinfo output i think you use the opencl cpu intel driver

[arpu]

for the GPU intel use beignet opencl driver

[psychocrypt]

@arpu Beignet OpenCl is not running with my kaby lake Intel.
I also thought that I am using my CPU but this screenshot shows that the CPU is not used.

% clinfo           
Number of platforms                               1
  Platform Name                                   Intel(R) OpenCL
  Platform Vendor                                 Intel(R) Corporation
  Platform Version                                OpenCL 2.0 
  Platform Profile                                FULL_PROFILE
  Platform Extensions                             cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_depth_images cl_khr_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_image2d_from_buffer cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_spir
  Platform Extensions function suffix             INTEL

  Platform Name                                   Intel(R) OpenCL
Number of devices                                 2
  Device Name                                     Intel(R) HD Graphics
  Device Vendor                                   Intel(R) Corporation
  Device Vendor ID                                0x8086
  Device Version                                  OpenCL 2.0 
  Driver Version                                  r4.1.61547
  Device OpenCL C Version                         OpenCL C 2.0 
  Device Type                                     GPU
  Device Profile                                  FULL_PROFILE
  Max compute units                               23
  Max clock frequency                             1050MHz
  Device Partition                                (core)
    Max number of sub-devices                     0
    Supported partition types                     by <unknown> (0x7FD900000000)
  Max work item dimensions                        3
  Max work item sizes                             256x256x256
  Max work group size                             256
  Preferred work group size multiple              32
  Preferred / native vector sizes                 
    char                                                16 / 16      
    short                                                8 / 8       
    int                                                  4 / 4       
    long                                                 1 / 1       
    half                                                 8 / 8        (cl_khr_fp16)
    float                                                1 / 1       
    double                                               1 / 1        (cl_khr_fp64)
  Half-precision Floating-point support           (cl_khr_fp16)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Single-precision Floating-point support         (core)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  Yes
  Double-precision Floating-point support         (cl_khr_fp64)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Address bits                                    64, Little-Endian
  Global memory size                              13254747751 (12.34GiB)
  Error Correction support                        No
  Max memory allocation                           4294959103 (4GiB)
  Unified memory for Host and Device              Yes
  Shared Virtual Memory (SVM) capabilities        (core)
    Coarse-grained buffer sharing                 Yes
    Fine-grained buffer sharing                   No
    Fine-grained system sharing                   No
    Atomics                                       No
  Minimum alignment for any data type             128 bytes
  Alignment of base address                       1024 bits (128 bytes)
  Preferred alignment for atomics                 
    SVM                                           64 bytes
    Global                                        64 bytes
    Local                                         64 bytes
  Max size for global variable                    65536 (64KiB)
  Preferred total size of global vars             4294959103 (4GiB)
  Global Memory cache type                        Read/Write
  Global Memory cache size                        524288
  Global Memory cache line                        64 bytes
  Image support                                   Yes
    Max number of samplers per kernel             16
    Max size for 1D images from buffer            268434943 pixels
    Max 1D or 2D image array size                 2048 images
    Base address alignment for 2D image buffers   4 bytes
    Pitch alignment for 2D image buffers          4 bytes
    Max 2D image size                             16384x16384 pixels
    Max 3D image size                             16384x16384x2048 pixels
    Max number of read image args                 128
    Max number of write image args                128
    Max number of read/write image args           128
  Max number of pipe args                         16
  Max active pipe reservations                    1
  Max pipe packet size                            1024
  Local memory type                               Local
  Local memory size                               65536 (64KiB)
  Max constant buffer size                        4294959103 (4GiB)
  Max number of constant args                     8
  Max size of kernel argument                     1024
  Queue properties (on host)                      
    Out-of-order execution                        Yes
    Profiling                                     Yes
  Queue properties (on device)                    
    Out-of-order execution                        Yes
    Profiling                                     Yes
    Preferred size                                131072 (128KiB)
    Max size                                      67108864 (64MiB)
  Max queues on device                            1
  Max events on device                            1024
  Prefer user sync for interop                    Yes
  Profiling timer resolution                      83ns
  Execution capabilities                          
    Run OpenCL kernels                            Yes
    Run native kernels                            No
    SPIR versions                                 1.2 
  printf() buffer size                            4194304 (4MiB)
  Built-in kernels                                block_motion_estimate_intel;block_advanced_motion_estimate_check_intel;block_advanced_motion_estimate_bidirectional_check_intel
  Motion Estimation accelerator version	(Intel)   2
  Device Available                                Yes
  Compiler Available                              Yes
  Linker Available                                Yes
  Device Extensions                               cl_intel_accelerator cl_intel_advanced_motion_estimation cl_intel_device_side_avc_motion_estimation cl_intel_driver_diagnostics cl_intel_media_block_io cl_intel_motion_estimation cl_intel_planar_yuv cl_intel_packed_yuv cl_intel_required_subgroup_size cl_intel_subgroups cl_intel_subgroups_short cl_intel_va_api_media_sharing cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_depth_images cl_khr_fp16 cl_khr_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_image2d_from_buffer cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_mipmap_image cl_khr_mipmap_image_writes cl_khr_spir cl_khr_subgroups 

  Device Name                                     Intel(R) Core(TM) i7-7500U CPU @ 2.70GHz
  Device Vendor                                   Intel(R) Corporation
  Device Vendor ID                                0x8086
  Device Version                                  OpenCL 2.0 (Build 400)
  Driver Version                                  1.2.0.400
  Device OpenCL C Version                         OpenCL C 2.0 
  Device Type                                     CPU
  Device Profile                                  FULL_PROFILE
  Max compute units                               4
  Max clock frequency                             2700MHz
  Device Partition                                (core)
    Max number of sub-devices                     4
    Supported partition types                     by counts, equally, by names (Intel)
  Max work item dimensions                        3
  Max work item sizes                             8192x8192x8192
  Max work group size                             8192
  Preferred work group size multiple              128
  Preferred / native vector sizes                 
    char                                                 1 / 32      
    short                                                1 / 16      
    int                                                  1 / 8       
    long                                                 1 / 4       
    half                                                 0 / 0        (n/a)
    float                                                1 / 8       
    double                                               1 / 4        (cl_khr_fp64)
  Half-precision Floating-point support           (n/a)
  Single-precision Floating-point support         (core)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 No
    Round to infinity                             No
    IEEE754-2008 fused multiply-add               No
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Double-precision Floating-point support         (cl_khr_fp64)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Address bits                                    64, Little-Endian
  Global memory size                              16578920448 (15.44GiB)
  Error Correction support                        No
  Max memory allocation                           4144730112 (3.86GiB)
  Unified memory for Host and Device              Yes
  Shared Virtual Memory (SVM) capabilities        (core)
    Coarse-grained buffer sharing                 Yes
    Fine-grained buffer sharing                   No
    Fine-grained system sharing                   No
    Atomics                                       No
  Minimum alignment for any data type             128 bytes
  Alignment of base address                       1024 bits (128 bytes)
  Preferred alignment for atomics                 
    SVM                                           64 bytes
    Global                                        64 bytes
    Local                                         0 bytes
  Max size for global variable                    65536 (64KiB)
  Preferred total size of global vars             65536 (64KiB)
  Global Memory cache type                        Read/Write
  Global Memory cache size                        262144
  Global Memory cache line                        64 bytes
  Image support                                   Yes
    Max number of samplers per kernel             480
    Max size for 1D images from buffer            259045632 pixels
    Max 1D or 2D image array size                 2048 images
    Base address alignment for 2D image buffers   64 bytes
    Pitch alignment for 2D image buffers          64 bytes
    Max 2D image size                             16384x16384 pixels
    Max 3D image size                             2048x2048x2048 pixels
    Max number of read image args                 480
    Max number of write image args                480
    Max number of read/write image args           480
  Max number of pipe args                         16
  Max active pipe reservations                    65535
  Max pipe packet size                            1024
  Local memory type                               Global
  Local memory size                               32768 (32KiB)
  Max constant buffer size                        131072 (128KiB)
  Max number of constant args                     480
  Max size of kernel argument                     3840 (3.75KiB)
  Queue properties (on host)                      
    Out-of-order execution                        Yes
    Profiling                                     Yes
    Local thread execution (Intel)                Yes
  Queue properties (on device)                    
    Out-of-order execution                        Yes
    Profiling                                     Yes
    Preferred size                                4294967295 (4GiB)
    Max size                                      4294967295 (4GiB)
  Max queues on device                            4294967295
  Max events on device                            4294967295
  Prefer user sync for interop                    No
  Profiling timer resolution                      1ns
  Execution capabilities                          
    Run OpenCL kernels                            Yes
    Run native kernels                            Yes
    SPIR versions                                 1.2
  printf() buffer size                            1048576 (1024KiB)
  Built-in kernels                                
  Device Available                                Yes
  Compiler Available                              Yes
  Linker Available                                Yes
  Device Extensions                               cl_khr_icd 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_byte_addressable_store cl_khr_depth_images cl_khr_3d_image_writes cl_intel_exec_by_local_thread cl_khr_spir cl_khr_fp64 cl_khr_image2d_from_buffer 

NULL platform behavior
  clGetPlatformInfo(NULL, CL_PLATFORM_NAME, ...)  No platform
  clGetDeviceIDs(NULL, CL_DEVICE_TYPE_ALL, ...)   No platform
  clCreateContext(NULL, ...) [default]            No platform
  clCreateContext(NULL, ...) [other]              Success [INTEL]
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CPU)  No platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_GPU)  No platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ACCELERATOR)  No platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CUSTOM)  No platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ALL)  No platform

[arpu]

ok you can look at intel_gpu_top tool to see some GPU top statistic
you can try build beignet from git but you need LLVM 3.9 for opencl 2.0

anyway i try to use this on my intel beignet installation right now for test

[psychocrypt]

intel_gpu_top       
(intel_gpu_top:5797) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file ../../lib/intel_mmio.c:145:
(intel_gpu_top:5797) intel-mmio-CRITICAL: Failed assertion: !(error != 0)
(intel_gpu_top:5797) intel-mmio-CRITICAL: Last errno: 13, Permission denied
(intel_gpu_top:5797) intel-mmio-CRITICAL: Couldn't map MMIO region
Stack trace:
  #0 [_init+0x83d9]
  #1 [_init+0x6036]
  #2 [_init+0x1301]
  #3 [__libc_start_main+0xf0]
  #4 [_init+0x2f61]
  #5 [<unknown>+0x2f61]
Test (null) failed.
**** DEBUG ****
(intel_gpu_top:5797) intel-mmio-CRITICAL: Test assertion failure function intel_mmio_use_pci_bar, file ../../lib/intel_mmio.c:145:
(intel_gpu_top:5797) intel-mmio-CRITICAL: Failed assertion: !(error != 0)
(intel_gpu_top:5797) intel-mmio-CRITICAL: Last errno: 13, Permission denied
(intel_gpu_top:5797) intel-mmio-CRITICAL: Couldn't map MMIO region
****  END  ****
FAIL (-1.000s)

I love Intel

update The tool needs root right (thx Intel).

[arpu]

sudo intel_gpu_top

[psychocrypt]

Yes I found it and I know why I am a Nvidia Fanboy :-).

result: user: 0.000000s, sys: 0.000000s, elapsed: 0.000256s, CPU: 0.0%, GPU: 100.0%

[arpu]

i tested with nvidea work fine! with opencl verion OpenCL 1.2 CUDA

[arpu]

i will try to do some more optimations

[psychocrypt]

@arpu But please only optimization for Intel or AMD. NVidia is not the target, we have a Nvidia version which is much faster than openCL.

Side note (why I am not focused on NVidia OpenCL):
I know that it is possible for NVidia GPUs to use use inline PTX code. I have also a OpenCL version where 4 threads working on 1 AES hash, but OpenCL is slower than CUDA.

[mk148a]

congratulations psychocrypt,
can i use on windows? I have 4790k cpu can i use it for hd graphics?

[dominikamann]

Hi,

with my Intel Card it also does not work:

[2017-07-07 10:14:51] : Compiling code and initializing GPUs. This will take a while...
[2017-07-07 10:14:51] : Device 0 work size 8 / 256.
[2017-07-07 10:14:51] : Error CL_BUILD_PROGRAM_FAILURE when calling clBuildProgram.
Build log:
1:16:26: warning: unknown OpenCL extension 'cl_amd_media_ops2' - ignoring
In file included from 1:18:
.\opencl/wolf-aes.cl:79:14: error: implicit declaration of function 'amd_bfe' is invalid in OpenCL
.\opencl/wolf-aes.cl:74:21: note: expanded from macro 'BYTE'
In file included from 1:19:
.\opencl/wolf-skein.cl:33:29: error: implicit declaration of function 'amd_bitalign' is invalid in OpenCL
header_with_defs_gpu.h:289:39: note: expanded from macro 'as_ulong'
In file included from 1:19:
.\opencl/wolf-skein.cl:33:20: error: invalid reinterpretation: sizes of 'ulong' (aka 'unsigned long') and 'int' must match
header_with_defs_gpu.h:289:21: note: expanded from macro 'as_ulong'
In file included from 1:19:
.\opencl/wolf-skein.cl:34:14: error: invalid reinterpretation: sizes of 'ulong' (aka 'unsigned long') and 'int' must match
header_with_defs_gpu.h:289:21: note: expanded from macro 'as_ulong'
1:328:26: warning: unknown OpenCL extension 'cl_amd_media_ops2' - ignoring
Press any key to exit.

Config:

"gpu_thread_num" : 1,
"gpu_threads_conf" : [
{ "index" : 0, "intensity" : 100, "worksize" : 8, "affine_to_cpu" : false },
],
"platform_index" : 0,

[psychocrypt]

@Kinimodnnama You need to use the dev branch. Do not forget to run make install else you will have the wrong opencl code in your folder.

[psychocrypt]

@mk148a I think currently not I need to fix cmake to support windows. After that it should work.

Note My intel gpu is slow and avoid the usage of a cpu miner during the gpu is mining.

[dominikamann]

@psychocrypt Thank you for the fast answer. Where can i download the binaries of this version? :)

[psychocrypt]

We have no binaries for the development version. You need to compile it by zour own. Warning we have currently no documentaion how to compile, if you are no geak you should wait for the next release or the documentation.

[dominikamann]

Okay thank you ;)

[soql]

Is somebody successful with OpenCL 1.2 ? I install intel opencl drivers but have all on "GPU COMPUTE ERROR" :-(

[snazzybunny]

A windows binary version would be awesome. Have an iris hd 580 to test with.

[JayWalker512]

Tested on Core M 5Y10

I was able to get the software (as of commit 7a4dc19) to run under Ubuntu 16.04 on Intel Core M 5Y10 with HD 5300 graphics, though I did run in to some issues.

I mostly followed the instructions from @psychocrypt above. First install the Intel OpenCL drivers which unfortunately requires a kernel rebuild around the Linux 4.7 kernel. I followed this guide and used the scripts therein: https://software.intel.com/en-us/articles/sdk-for-opencl-gsg

I also installed the Intel OpenCL SDK, though I am not sure if this is strictly necessary, the Intel guide just instructed to do so. You have to jump through some hoops to download: https://software.intel.com/en-us/intel-opencl/download

I was able to get the software to at least start with this configuration:

"gpu_thread_num" : 1,

"gpu_threads_conf" : [ 
	{ "index" : 0, "intensity" : 32, "worksize" : 8, "affine_to_cpu" : false },
],

"platform_index" : 0,

Though this is probably not ideal, it would crash with many other configurations (includng the default), though it seemed to be working with this at ~20 H/s. Could you post what config you used above @psychocrypt?

Worth noting, the software would occasionally give the error:

[447.121859] igfxdcd: failed to locate the system routine

Not sure of the cause, but it didn't seem to impact the program running.

For reference, here is some data about my system:

$ uname -a
Linux jaywalker-UX305FA 4.7.0.intel.r5.0 #1 SMP Mon Aug 21 22:20:34 EDT 2017 x86_64 x86_64 x86_64 GNU/Linux

$ clinfo
Number of platforms                               2
  Platform Name                                   Intel(R) OpenCL
  Platform Vendor                                 Intel(R) Corporation
  Platform Version                                OpenCL 2.0 
  Platform Profile                                FULL_PROFILE
  Platform Extensions                             cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_depth_images cl_khr_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_image2d_from_buffer cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_spir
  Platform Extensions function suffix             INTEL

  Platform Name                                   Experimental OpenCL 2.1 CPU Only Platform
  Platform Vendor                                 Intel(R) Corporation
  Platform Version                                OpenCL 2.1 LINUX
  Platform Profile                                FULL_PROFILE
  Platform Extensions                             cl_khr_icd 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_byte_addressable_store cl_khr_depth_images cl_khr_3d_image_writes cl_intel_exec_by_local_thread cl_khr_spir cl_khr_fp64 cl_khr_image2d_from_buffer 
  Platform Host timer resolution                  1ns
  Platform Extensions function suffix             INTEL

  Platform Name                                   Intel(R) OpenCL
Number of devices                                 2
  Device Name                                     Intel(R) HD Graphics
  Device Vendor                                   Intel(R) Corporation
  Device Vendor ID                                0x8086
  Device Version                                  OpenCL 2.0 
  Driver Version                                  r5.0.63503
  Device OpenCL C Version                         OpenCL C 2.0 
  Device Type                                     GPU
  Device Profile                                  FULL_PROFILE
  Max compute units                               24
  Max clock frequency                             800MHz
  Device Partition                                (core)
    Max number of sub-devices                     0
    Supported partition types                     None
  Max work item dimensions                        3
  Max work item sizes                             256x256x256
  Max work group size                             256
  Preferred work group size multiple              32
  Preferred / native vector sizes                 
    char                                                16 / 16      
    short                                                8 / 8       
    int                                                  4 / 4       
    long                                                 1 / 1       
    half                                                 8 / 8        (cl_khr_fp16)
    float                                                1 / 1       
    double                                               1 / 1        (cl_khr_fp64)
  Half-precision Floating-point support           (cl_khr_fp16)
    Denormals                                     No
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Single-precision Floating-point support         (core)
    Denormals                                     No
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  Yes
  Double-precision Floating-point support         (cl_khr_fp64)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Address bits                                    64, Little-Endian
  Global memory size                              6210191360 (5.784GiB)
  Error Correction support                        No
  Max memory allocation                           3105095680 (2.892GiB)
  Unified memory for Host and Device              Yes
  Shared Virtual Memory (SVM) capabilities        (core)
    Coarse-grained buffer sharing                 Yes
    Fine-grained buffer sharing                   No
    Fine-grained system sharing                   No
    Atomics                                       No
  Minimum alignment for any data type             128 bytes
  Alignment of base address                       1024 bits (128 bytes)
  Preferred alignment for atomics                 
    SVM                                           64 bytes
    Global                                        64 bytes
    Local                                         64 bytes
  Max size for global variable                    65536 (64KiB)
  Preferred total size of global vars             3105095680 (2.892GiB)
  Global Memory cache type                        Read/Write
  Global Memory cache size                        589824
  Global Memory cache line                        64 bytes
  Image support                                   Yes
    Max number of samplers per kernel             16
    Max size for 1D images from buffer            194068480 pixels
    Max 1D or 2D image array size                 2048 images
    Base address alignment for 2D image buffers   4 bytes
    Pitch alignment for 2D image buffers          4 bytes
    Max 2D image size                             16384x16384 pixels
    Max 3D image size                             2048x2048x2048 pixels
    Max number of read image args                 128
    Max number of write image args                128
    Max number of read/write image args           128
  Max number of pipe args                         16
  Max active pipe reservations                    1
  Max pipe packet size                            1024
  Local memory type                               Local
  Local memory size                               65536 (64KiB)
  Max constant buffer size                        3105095680 (2.892GiB)
  Max number of constant args                     8
  Max size of kernel argument                     1024
  Queue properties (on host)                      
    Out-of-order execution                        Yes
    Profiling                                     Yes
  Queue properties (on device)                    
    Out-of-order execution                        Yes
    Profiling                                     Yes
    Preferred size                                131072 (128KiB)
    Max size                                      67108864 (64MiB)
  Max queues on device                            1
  Max events on device                            1024
  Prefer user sync for interop                    Yes
  Profiling timer resolution                      80ns
  Execution capabilities                          
    Run OpenCL kernels                            Yes
    Run native kernels                            No
    SPIR versions                                 1.2 
  printf() buffer size                            4194304 (4MiB)
  Built-in kernels                                block_motion_estimate_intel;block_advanced_motion_estimate_check_intel;block_advanced_motion_estimate_bidirectional_check_intel
  Motion Estimation accelerator version	(Intel)   2
  Device Available                                Yes
  Compiler Available                              Yes
  Linker Available                                Yes
  Device Extensions                               cl_intel_accelerator cl_intel_advanced_motion_estimation cl_intel_device_side_avc_motion_estimation cl_intel_driver_diagnostics cl_intel_media_block_io cl_intel_motion_estimation cl_intel_planar_yuv cl_intel_packed_yuv cl_intel_required_subgroup_size cl_intel_subgroups cl_intel_va_api_media_sharing cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_depth_images cl_khr_fp16 cl_khr_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_image2d_from_buffer cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_mipmap_image cl_khr_mipmap_image_writes cl_khr_spir 

  Device Name                                     Intel(R) Core(TM) M-5Y10c CPU @ 0.80GHz
  Device Vendor                                   Intel(R) Corporation
  Device Vendor ID                                0x8086
  Device Version                                  OpenCL 2.0 (Build 475)
  Driver Version                                  1.2.0.475
  Device OpenCL C Version                         OpenCL C 2.0 
  Device Type                                     CPU
  Device Profile                                  FULL_PROFILE
  Max compute units                               4
  Max clock frequency                             800MHz
  Device Partition                                (core)
    Max number of sub-devices                     4
    Supported partition types                     by counts, equally, by names (Intel)
  Max work item dimensions                        3
  Max work item sizes                             8192x8192x8192
  Max work group size                             8192
  Preferred work group size multiple              128
  Preferred / native vector sizes                 
    char                                                 1 / 32      
    short                                                1 / 16      
    int                                                  1 / 8       
    long                                                 1 / 4       
    half                                                 0 / 0        (n/a)
    float                                                1 / 8       
    double                                               1 / 4        (cl_khr_fp64)
  Half-precision Floating-point support           (n/a)
  Single-precision Floating-point support         (core)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 No
    Round to infinity                             No
    IEEE754-2008 fused multiply-add               No
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Double-precision Floating-point support         (cl_khr_fp64)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Address bits                                    64, Little-Endian
  Global memory size                              7773224960 (7.239GiB)
  Error Correction support                        No
  Max memory allocation                           1943306240 (1.81GiB)
  Unified memory for Host and Device              Yes
  Shared Virtual Memory (SVM) capabilities        (core)
    Coarse-grained buffer sharing                 Yes
    Fine-grained buffer sharing                   No
    Fine-grained system sharing                   No
    Atomics                                       No
  Minimum alignment for any data type             128 bytes
  Alignment of base address                       1024 bits (128 bytes)
  Preferred alignment for atomics                 
    SVM                                           64 bytes
    Global                                        64 bytes
    Local                                         0 bytes
  Max size for global variable                    65536 (64KiB)
  Preferred total size of global vars             65536 (64KiB)
  Global Memory cache type                        Read/Write
  Global Memory cache size                        262144
  Global Memory cache line                        64 bytes
  Image support                                   Yes
    Max number of samplers per kernel             480
    Max size for 1D images from buffer            121456640 pixels
    Max 1D or 2D image array size                 2048 images
    Base address alignment for 2D image buffers   64 bytes
    Pitch alignment for 2D image buffers          64 bytes
    Max 2D image size                             16384x16384 pixels
    Max 3D image size                             2048x2048x2048 pixels
    Max number of read image args                 480
    Max number of write image args                480
    Max number of read/write image args           480
  Max number of pipe args                         16
  Max active pipe reservations                    65535
  Max pipe packet size                            1024
  Local memory type                               Global
  Local memory size                               32768 (32KiB)
  Max constant buffer size                        131072 (128KiB)
  Max number of constant args                     480
  Max size of kernel argument                     3840 (3.75KiB)
  Queue properties (on host)                      
    Out-of-order execution                        Yes
    Profiling                                     Yes
    Local thread execution (Intel)                Yes
  Queue properties (on device)                    
    Out-of-order execution                        Yes
    Profiling                                     Yes
    Preferred size                                4294967295 (4GiB)
    Max size                                      4294967295 (4GiB)
  Max queues on device                            4294967295
  Max events on device                            4294967295
  Prefer user sync for interop                    No
  Profiling timer resolution                      1ns
  Execution capabilities                          
    Run OpenCL kernels                            Yes
    Run native kernels                            Yes
    SPIR versions                                 1.2
  printf() buffer size                            1048576 (1024KiB)
  Built-in kernels                                
  Device Available                                Yes
  Compiler Available                              Yes
  Linker Available                                Yes
  Device Extensions                               cl_khr_icd 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_byte_addressable_store cl_khr_depth_images cl_khr_3d_image_writes cl_intel_exec_by_local_thread cl_khr_spir cl_khr_fp64 cl_khr_image2d_from_buffer 

  Platform Name                                   Experimental OpenCL 2.1 CPU Only Platform
Number of devices                                 1
  Device Name                                     Intel(R) Core(TM) M-5Y10c CPU @ 0.80GHz
  Device Vendor                                   Intel(R) Corporation
  Device Vendor ID                                0x8086
  Device Version                                  OpenCL 2.1 (Build 10)
  Driver Version                                  1.2.0.10
  Device OpenCL C Version                         OpenCL C 2.0 
  Device Type                                     CPU
  Device Profile                                  FULL_PROFILE
  Max compute units                               4
  Max clock frequency                             800MHz
  Device Partition                                (core)
    Max number of sub-devices                     4
    Supported partition types                     by counts, equally, by names (Intel)
  Max work item dimensions                        3
  Max work item sizes                             8192x8192x8192
  Max work group size                             8192
  Preferred work group size multiple              128
  Max sub-groups per work group                   1
  Preferred / native vector sizes                 
    char                                                 1 / 32      
    short                                                1 / 16      
    int                                                  1 / 8       
    long                                                 1 / 4       
    half                                                 0 / 0        (n/a)
    float                                                1 / 8       
    double                                               1 / 4        (cl_khr_fp64)
  Half-precision Floating-point support           (n/a)
  Single-precision Floating-point support         (core)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 No
    Round to infinity                             No
    IEEE754-2008 fused multiply-add               No
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Double-precision Floating-point support         (cl_khr_fp64)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Address bits                                    64, Little-Endian
  Global memory size                              7773224960 (7.239GiB)
  Error Correction support                        No
  Max memory allocation                           1943306240 (1.81GiB)
  Unified memory for Host and Device              Yes
  Shared Virtual Memory (SVM) capabilities        (core)
    Coarse-grained buffer sharing                 Yes
    Fine-grained buffer sharing                   Yes
    Fine-grained system sharing                   Yes
    Atomics                                       Yes
  Minimum alignment for any data type             128 bytes
  Alignment of base address                       1024 bits (128 bytes)
  Preferred alignment for atomics                 
    SVM                                           64 bytes
    Global                                        64 bytes
    Local                                         0 bytes
  Max size for global variable                    65536 (64KiB)
  Preferred total size of global vars             65536 (64KiB)
  Global Memory cache type                        Read/Write
  Global Memory cache size                        262144
  Global Memory cache line                        64 bytes
  Image support                                   Yes
    Max number of samplers per kernel             480
    Max size for 1D images from buffer            121456640 pixels
    Max 1D or 2D image array size                 2048 images
    Base address alignment for 2D image buffers   64 bytes
    Pitch alignment for 2D image buffers          64 bytes
    Max 2D image size                             16384x16384 pixels
    Max 3D image size                             2048x2048x2048 pixels
    Max number of read image args                 480
    Max number of write image args                480
    Max number of read/write image args           480
  Max number of pipe args                         16
  Max active pipe reservations                    65535
  Max pipe packet size                            1024
  Local memory type                               Global
  Local memory size                               32768 (32KiB)
  Max constant buffer size                        131072 (128KiB)
  Max number of constant args                     480
  Max size of kernel argument                     3840 (3.75KiB)
  Queue properties (on host)                      
    Out-of-order execution                        Yes
    Profiling                                     Yes
    Local thread execution (Intel)                Yes
  Queue properties (on device)                    
    Out-of-order execution                        Yes
    Profiling                                     Yes
    Preferred size                                4294967295 (4GiB)
    Max size                                      4294967295 (4GiB)
  Max queues on device                            4294967295
  Max events on device                            4294967295
  Prefer user sync for interop                    No
  Profiling timer resolution                      1ns
  Execution capabilities                          
    Run OpenCL kernels                            Yes
    Run native kernels                            Yes
    Sub-group independent forward progress        No
    IL version                                    SPIR-V_1.0
    SPIR versions                                 1.2
  printf() buffer size                            1048576 (1024KiB)
  Built-in kernels                                
  Device Available                                Yes
  Compiler Available                              Yes
  Linker Available                                Yes
  Device Extensions                               cl_khr_icd 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_byte_addressable_store cl_khr_depth_images cl_khr_3d_image_writes cl_intel_exec_by_local_thread cl_khr_spir cl_khr_fp64 cl_khr_image2d_from_buffer 

NULL platform behavior
  clGetPlatformInfo(NULL, CL_PLATFORM_NAME, ...)  No platform
  clGetDeviceIDs(NULL, CL_DEVICE_TYPE_ALL, ...)   No platform
  clCreateContext(NULL, ...) [default]            No platform
  clCreateContext(NULL, ...) [other]              Success [INTEL]
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CPU)  No platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_GPU)  No platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ACCELERATOR)  No platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CUSTOM)  No platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ALL)  No platform

Intel provides a system analyzer utility python script https://software.intel.com/en-us/articles/mss-sys-analyzer-linux
This is my output:

./sys_analyzer_linux.py 
--------------------------
Hardware readiness checks:
--------------------------
 [ OK ] Processor name: Intel(R) Core(TM) M-5Y10c CPU @ 0.80GHz
--------------------------
OS readiness checks:
--------------------------
 [ OK ] GPU visible to OS
--------------------------
Media Server Studio Install:
--------------------------
 [ OK ] user in video group
 [ ERROR ] libva.so.1 not found.  Check LD_LIBRARY_PATH contains '/usr/lib64;/usr/local/lib'
 [ ERROR ] libva not loading Intel iHD
 [ ERROR ] vainfo not reporting codec entry points
 [ ERROR ] no libva include files.  Are Intel components installed?
--------------------------
Component Smoke Tests:
--------------------------
 [ ERROR ] no Media SDK include files.  Are Intel components installed?
 [ OK ] OpenCL check:platform:Intel(R) OpenCL GPU OK CPU OK
platform:Experimental OpenCL 2.1 CPU Only Platform GPU OK CPU OK

If I get time maybe I will try this with my desktop i5 4440 and Intel HD 4600. It probably will require different steps as the Intel driver package above says it is for 5th gen and above only (Haswell is 4th gen IIRC). I also don't think it supports above OpenCL 1.2.

Also re: CPU performance @psychocrypt while running the GPU + CPU miner on Intel HD, I believe the IGP shares some L3 cache with the CPU (unclear if it is a reserved amount), so perhaps the performance drop is related to contention there somehow. Have you tried running just one CPU thread and just one GPU thread to see if they affect each other?

Also worth noting about performance: when running the miner my GUI environment was basically unusable, everything was incredibly unresponsive (but I guess that's not surprising when smacking the GPU with work!). Did you experience this @psychocrypt ?

[chris-pang]

I tried compile it with intel-opencl-r5 and manage to run it with average 108 h/s.
This is without CPU miner and htop showing 0% for all 8 CPU thread. However, I noticed that after a couple of shares found, it immediately hogging 2 CPU thread 100% all the time.

I did a few tries, and every time it hogs 200%/800% after couples of shares/minutes(like after 5 minutes).

Platform:
Centos 7.3
Intel(R) Core(TM) i7-6700K CPU @ 4.00GHz

[naortega]

I've been trying to run this on my computer with an Intel(R) Core(TM) i3-5005U CPU @ 2.00GHz CPU and Intel(R) HD Graphics GPU, and I made sure to work with the dev branch. When running the program I get the following error:

nicolas:bin/ (dev✗) $ ./xmr-stak-amd config.txt                                                                                                                 [0:43:56]
[2017-09-10 00:44:17] : Compiling code and initializing GPUs. This will take a while...
[2017-09-10 00:44:17] : Error CL_DEVICE_NOT_FOUND when calling clGetDeviceIDs for number of devices.

Here's a link to my clinfo output: http://paste.debian.net/985329/

According to the API it should only give me this error if no devices of the specified device_type (which in this case is CL_DEVICE_TYPE_GPU) are found.

Follows is a section of the config.txt:

/* 
 * Number of GPUs that you have in your system. Each GPU will get its own CPU thread.
 */
"gpu_thread_num" : 1,

/*
 * GPU configuration. You should play around with intensity and worksize as the fastest settings will vary.
 *      index    - GPU index number usually starts from 0
 *  intensity    - Number of parallel GPU threads (nothing to do with CPU threads)
 *   worksize    - Number of local GPU threads (nothing to do with CPU threads)
 * affine_to_cpu - This will affine the thread to a CPU. This can make a GPU miner play along nicer with a CPU miner.
 */
"gpu_threads_conf" : [ 
        { "index" : 0, "intensity" : 23, "worksize" : 8, "affine_to_cpu" : false },
],

/*
 * Platform index. This will be 0 unless you have different OpenCL platform - eg. AMD and Intel.
 */
"platform_index" : 0,

[psychocrypt]

@Deathsbreed Please try to set index to 1 else you are mining with your cpu.

[naortega]

Alright, thanks. That along with some other modifications to the config.txt (otherwise I was getting an error about CL workgroup size, but it seems to work at "worksize" : 4) helped to get it running. Though it definitely doesn't seem worth it with my GPU running at 30H/s + slower CPU mining. :P

Thanks for the help.

[nikl11]

Hello, trying to run the miner under Windows 10 on Intel GPU. I have installed the original Intel driver (Intel VGA driver 20.19.15.4463) and Intel SDK opencl 2.0. I managed to compile the source codes from the dev branch (the binary is running fine on my vega 56 on my other machine). However, I cannot make it run on my integrated GPU, the output is

[2017-09-26 03:09:32] : Compiling code and initializing GPUs. This will take a while...
[2017-09-26 03:09:33] : Device 0 work size 8 / 256.
[2017-09-26 03:09:33] : Error CL_BUILD_PROGRAM_FAILURE when calling clBuildProgram.
Build log:
1:16:26: warning: unknown OpenCL extension 'cl_amd_media_ops2' - ignoring
In file included from 1:18:
.\opencl/wolf-aes.cl:79:14: error: implicit declaration of function 'amd_bfe' is invalid in OpenCL
.\opencl/wolf-aes.cl:74:21: note: expanded from macro 'BYTE'
In file included from 1:19:
.\opencl/wolf-skein.cl:33:29: error: implicit declaration of function 'amd_bitalign' is invalid in OpenCL header_with_defs_gpu.h:289:39: note: expanded from macro 'as_ulong'
In file included from 1:19:
.\opencl/wolf-skein.cl:33:20: error: invalid reinterpretation: sizes of 'ulong' (aka 'unsigned long') and 'int' must match header_with_defs_gpu.h:289:21: note: expanded from macro 'as_ulong'
In file included from 1:19:
.\opencl/wolf-skein.cl:34:14: error: invalid reinterpretation: sizes of 'ulong' (aka 'unsigned long') and 'int' must match header_with_defs_gpu.h:289:21: note: expanded from macro 'as_ulong'
1:328:26: warning: unknown OpenCL extension 'cl_amd_media_ops2' - ignoring
Press any key to exit.

Not sure how to run clinfo on windows, but ethminer -G --list-devices outputs

Listing OpenCL devices.
FORMAT: [platformID] [deviceID] deviceName
[0] [0] Intel(R) HD Graphics 5500
CL_DEVICE_TYPE: GPU
CL_DEVICE_GLOBAL_MEM_SIZE: 3391448679
CL_DEVICE_MAX_MEM_ALLOC_SIZE: 847862169
CL_DEVICE_MAX_WORK_GROUP_SIZE: 256
[1] [0] GeForce 920M
CL_DEVICE_TYPE: GPU
CL_DEVICE_GLOBAL_MEM_SIZE: 2147483648
CL_DEVICE_MAX_MEM_ALLOC_SIZE: 536870912
CL_DEVICE_MAX_WORK_GROUP_SIZE: 1024

and the main part of my config is

"gpu_thread_num" : 1,
"gpu_threads_conf" : [
{ "index" : 0, "intensity" : 64, "worksize" : 8, "affine_to_cpu" : false },
],
"platform_index" : 0,

Thanks for help @psychocrypt @Deathsbreed .

[JayWalker512]

@nikl11 just be sure, did you compile from the dev branch? That's where @psychocrypt did the work for non-AMD media extension GPU's and I don't think it's merged in to master yet. Your errors look like the miner is trying to use AMD extensions which obviously the Intel GPU doesn't have.

[nikl11]

Yes I did use the dev branch, but in the opencl folder I can see the source codes still contain amd extensions. I copy the opencl folder next to my binary.

EDIT: Ok, you were right, I was using an old cryptonight.cl file, after manual update it works fine. Only 20H/s tho :-). Sorry and thanks.
EDIT2: On i5-5200U's gpu it works, but on i5-3570k's it doesn't, it starts normally, looks like it is working, and after the message "Stratum found a new block", the terminal just exits, no error message.

[zomborg21]

I downloaded the dev branch and try to compile and I get an error.

In the logs it says

Build FAILED.
"C:\Users\XXXX\Desktop\GPU miner\xmr-stak-amd-dev\CMakeFiles\CMakeTmp\cmTC_f12cc.vcxproj" (default target) (1) ->

(ClCompile target) ->

C:\Users\XXX\Desktop\GPU miner\xmr-stak-amd-dev\CMakeFiles\CMakeTmp\CheckIncludeFile.c(1): fatal error C1083: Cannot open include file: 'pthread.h': No such file or directory [C:\Users\XXX\Desktop\GPU miner\xmr-stak-amd-dev\CMakeFiles\CMakeTmp\cmTC_f12cc.vcxproj]
0 Warning(s)
1 Error(s)
Time Elapsed 00:00:00.73

Any help would be appreciated I am Windows 10 and have an Intel GPU

[psychocrypt]

The next release within a few days will bring full windows kompile support. please do not wast time in intel gpus they are slower than slow.

[maknij]

Even though Intel GPU's are slow, are there any plans to support them in upcoming unified miner releases for Windows in any case? (they might be slow but every bit of hash power helps). Thanks

[psychocrypt]

Intel gpus are arlready supported but you need to write a amd.txt file where the intel platform is selected and configure the dive by hand. Trust me the gpu is slower than your cpu.

[maknij]

Thanks, just gave it a shot and see what you mean - CPU hashrate was also reduced when the GPU was being used. Will stick with CPU, thanks.

[psychocrypt]

maybe it will be better this year when intel is releasing the cpu with an amd gpu with hbm2 dedicated memory.

[jgillich]

I just tried xmr-stak with my Intel laptop, without success: https://gist.github.com/jgillich/3fb2982fd385008848e053b3cad9ca3a

Would be nice to get this working for development purposes. Seems like xmr-stak still runs some AMD specific instructions like amd_bfe and amd_bitalign. Or do I need the Intel OpenCL driver instead of beignet?

[JayWalker512]

@jgillich What hardware is in your laptop?
I have tried to run xmr-stak on HD 4600 with Beignet but it fails due to bugs in Beignet. I was able to get it running on an HD 5300 though (which supports OpenCL 2.0), but that required a number of steps that I listed above in this thread (see: Core M 5Y10). It's not easy in any case.

[jgillich]

I have a HD 620 and a Skylake i5. According to clinfo, OpenCL 2.0 is supported. I will have to try the Intel OpenCL driver, but it will probably be a pain to install on Solus.

[psychocrypt]

I think you used the wrong platform. You must use the intel opencl platform. As I remember right I added the instruction how to install the intel opencl libraries into the pull request linked within this issue.

[jgillich]

Seems to be that way, it works fine with the Intel libraries. Thanks!

[CybotDNA]

I have "xmr-stak-rel-2.1.0-Win64-XMR-2" installed.
Is this update in there already? How can I use my igpu? I have a 6600

[Rybec]

I am interested to know if this has made any progress. It has been a few years, and Intel has released new processors with more advanced GPUs. Has anyone tried this with more recent Intel GPUs, and are the CPUs still outperforming the GPUs?