Unable to activate OpenCL

ΩElias · October 15, 2019, 9:19am

Hello,
I have been trying to get LibreOffice 6.2 to use OpenCL to accelerate processing large spreadsheets (over 3MB raw data). I am running Ubuntu Mate 18.04.3 with the 5.3.1 Mainline kernel.
clinfo output:


    Number of platforms                               1
  Platform Name                                   Intel Gen OCL Driver
  Platform Vendor                                 Intel
  Platform Version                                OpenCL 2.0 beignet 1.3
  Platform Profile                                FULL_PROFILE
  Platform Extensions                             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_3d_image_writes cl_khr_image2d_from_buffer cl_khr_depth_images cl_khr_spir cl_khr_icd cl_intel_accelerator cl_intel_subgroups cl_intel_subgroups_short cl_khr_gl_sharing
  Platform Extensions function suffix             Intel

  Platform Name                                   Intel Gen OCL Driver
Number of devices                                 1
  Device Name                                     Intel(R) HD Graphics 5500 BroadWell U-Processor GT2
  Device Vendor                                   Intel
  Device Vendor ID                                0x8086
  Device Version                                  OpenCL 1.2 beignet 1.3
  Driver Version                                  1.3
  Device OpenCL C Version                         OpenCL C 1.2 beignet 1.3
  Device Type                                     GPU
  Device Profile                                  FULL_PROFILE
  Device Available                                Yes
  Compiler Available                              Yes
  Linker Available                                Yes
  Max compute units                               24
  Max clock frequency                             1000MHz
  Device Partition                                (core)
    Max number of sub-devices                     1
    Supported partition types                     None, None, None
  Max work item dimensions                        3
  Max work item sizes                             512x512x512
  Max work group size                             512
  Preferred work group size multiple              16
  Preferred / native vector sizes                 
    char                                                16 / 8       
    short                                                8 / 8       
    int                                                  4 / 4       
    long                                                 2 / 2       
    half                                                 0 / 8        (cl_khr_fp16)
    float                                                4 / 4       
    double                                               0 / 2        (n/a)
  Half-precision Floating-point support           (cl_khr_fp16)
    Denormals                                     No
    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
  Single-precision Floating-point support         (core)
    Denormals                                     No
    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         (n/a)
  Address bits                                    32, Little-Endian
  Global memory size                              4294967296 (4GiB)
  Error Correction support                        No
  Max memory allocation                           2147483648 (2GiB)
  Unified memory for Host and Device              Yes
  Minimum alignment for any data type             128 bytes
  Alignment of base address                       1024 bits (128 bytes)
  Global Memory cache type                        Read/Write
  Global Memory cache size                        8192 (8KiB)
  Global Memory cache line size                   64 bytes
  Image support                                   Yes
    Max number of samplers per kernel             16
    Max size for 1D images from buffer            65536 pixels
    Max 1D or 2D image array size                 2048 images
    Base address alignment for 2D image buffers   4096 bytes
    Pitch alignment for 2D image buffers          1 pixels
    Max 2D image size                             8192x8192 pixels
    Max 3D image size                             8192x8192x2048 pixels
    Max number of read image args                 128
    Max number of write image args                8
  Local memory type                               Local
  Local memory size                               65536 (64KiB)
  Max number of constant args                     8
  Max constant buffer size                        134217728 (128MiB)
  Max size of kernel argument                     1024
  Queue properties                                
    Out-of-order execution                        No
    Profiling                                     Yes
  Prefer user sync for interop                    Yes
  Profiling timer resolution                      80ns
  Execution capabilities                          
    Run OpenCL kernels                            Yes
    Run native kernels                            Yes
    SPIR versions                                 1.2
  printf() buffer size                            1048576 (1024KiB)
  Built-in kernels                                __cl_copy_region_align4;__cl_copy_region_align16;__cl_cpy_region_unalign_same_offset;__cl_copy_region_unalign_dst_offset;__cl_copy_region_unalign_src_offset;__cl_copy_buffer_rect;__cl_copy_image_1d_to_1d;__cl_copy_image_2d_to_2d;__cl_copy_image_3d_to_2d;__cl_copy_image_2d_to_3d;__cl_copy_image_3d_to_3d;__cl_copy_image_2d_to_buffer;__cl_copy_image_3d_to_buffer;__cl_copy_buffer_to_image_2d;__cl_copy_buffer_to_image_3d;__cl_fill_region_unalign;__cl_fill_region_align2;__cl_fill_region_align4;__cl_fill_region_align8_2;__cl_fill_region_align8_4;__cl_fill_region_align8_8;__cl_fill_region_align8_16;__cl_fill_region_align128;__cl_fill_image_1d;__cl_fill_image_1d_array;__cl_fill_image_2d;__cl_fill_image_2d_array;__cl_fill_image_3d;
  Device Extensions                               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_3d_image_writes cl_khr_image2d_from_buffer cl_khr_depth_images cl_khr_spir cl_khr_icd cl_intel_accelerator cl_intel_subgroups cl_intel_subgroups_short cl_khr_gl_sharing cl_khr_fp16

NULL platform behavior
  clGetPlatformInfo(NULL, CL_PLATFORM_NAME, ...)  Intel Gen OCL Driver
  clGetDeviceIDs(NULL, CL_DEVICE_TYPE_ALL, ...)   Success [Intel]
  clCreateContext(NULL, ...) [default]            Success [Intel]
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_DEFAULT)  Success (1)
    Platform Name                                 Intel Gen OCL Driver
    Device Name                                   Intel(R) HD Graphics 5500 BroadWell U-Processor GT2
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CPU)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_GPU)  Success (1)
    Platform Name                                 Intel Gen OCL Driver
    Device Name                                   Intel(R) HD Graphics 5500 BroadWell U-Processor GT2
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ACCELERATOR)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CUSTOM)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ALL)  Success (1)
    Platform Name                                 Intel Gen OCL Driver
    Device Name                                   Intel(R) HD Graphics 5500 BroadWell U-Processor GT2

ICD loader properties
  ICD loader Name                                 OpenCL ICD Loader
  ICD loader Vendor                               OCL Icd free software
  ICD loader Version                              2.2.11
  ICD loader Profile                              OpenCL 2.1

Everytime I try to Select the OpenCL option it asks me to restart LibO. When it does the option is disabled again.
How can I fix this issue?

erAck · October 15, 2019, 11:59am

Unfortunately OpenCL implementations greatly vary between drivers and devices, and not all are suitable. Or rather, few are. In this case I guess it’s because it is not whitelisted. Or otherwise the driver doesn’t pass the tests LibreOffice carries out so it gets disabled. I’m wondering though that the Vendor is given just as Intel instead of Intel(R) Corporation (which then would match the whitelist).

You can try to add the device/driver details to your user profile’s registrymodifications.xcu (see https://wiki.documentfoundation.org/UserProfile#GNU.2FLinux for where it’s located) in OpenCLWhiteList and see if it works; note it may crash LibreOffice or even worse, calculate some things wrongly. Look for

<item oor:path="/org.openoffice.Office.Common/Misc"><prop oor:name="OpenCLWhiteList" oor:op="fuse">

(that should be present if you tried to enable OpenCL) and in the <value> list add an item

<it>Linux//Intel//OpenCL 2\.0 beignet 1\.3</it>

Note the // double slashes that delimit empty fields. I’m not entirely sure about the OpenCL 2\.0 beignet 1\.3 that should be the device driver version, it could also be just 1\.3 (note the string is a regular expression, hence the \. escaped dot).

I’m not responsible if this toasts your machine!

Please give feedback, if it works for you and calculates things correctly then we maybe could add the driver to the known whitelist.

ΩElias · October 15, 2019, 12:34pm

The list item above didn’t work. I’m gonna try messing around with this a little.
Btw. So far I’ve only managed to toast a GPU due to attempting to control a motor using the DDC interface

Edit: I’ve just found this string in here:

Intel(R) HD Graphics 5500 BroadWell U-Processor GT2

Edit2: Nvm, I remember adding this myself. derp

ΩElias · October 15, 2019, 12:55pm

Ok, So far I’ve tried 2.0 beignet 1.3, 1.3, 1.2 etc.
Is there a way to get a verbose log?

erAck · October 15, 2019, 5:49pm

Only in a debug build you’d have to compile yourself, it could output the detected device info (might be worth changing that to also production build). However, if it fails with all version strings tried then possibly it fails during the runtime check, in which case nothing would help.

Francewhoa · June 2, 2021, 12:29am

Hello @ΩElias I suggest running clinfo command to get the present status of your OpenCL. If the result is 0 (zero). Then you need to install the appropriate OpenCL package(s) for your CPU.

Below is the same as above. But with details if you’re interested in those.

Steps

Install this clinfo package Debian -- Package Search Results -- clinfo
If not already done, install this sudo package Debian -- Package Search Results -- sudo
Using Terminal, run this command get the present status of your OpenCL sudo clinfo
If Terminal return this message Number of platforms 0 Then it is likely that OpenCL is not install, or not configure properly, or the wrong package was installed.
If OpenCL is not install. You need to install the appropriate OpenCL package(s) for your CPU. For example, for Debian 10 Buster, with an Intel i7 CPU & NVIDIA graphics, install this package nvidia-opencl-icd at Debian -- Details of package nvidia-opencl-icd in buster
- Alternatively, a more recent version of this package is available in the Debian Backport at Debian -- Details of package nvidia-opencl-icd in buster-backports
Restart your device
Using Terminal, run this command again to get the present status of your OpenCL sudo clinfo. If the result is not zero then you have successfully installed OpenCL for your device.
Using LibreOffice, try to activate OpenCL again. Enjoy much faster calculations in LibreOffice.

spearthistle · February 3, 2023, 1:00pm

Hi Francewhoa,
It is 2023 and I have installed version 7.4.5 on Manjaro using snap.
I have installed

opencl-headers
opencl-icd-loader
intel-compute-runtime

and rebooted the machine.

I cannot get

Allow use of OpenCL

to activate (be ticked after the application restart).

clinfo

reports

Number of platforms Platform Name Platform Vendor Platform Version Platform Profile Platform Extensions Platform Extensions with Version cl_khr_device_uuid cl_khr_fp16 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_intel_command_queue_families cl_intel_subgroups cl_intel_required_subgroup_size cl_intel_subgroups_short cl_khr_spir cl_intel_accelerator cl_intel_driver_diagnostics cl_khr_priority_hints cl_khr_throttle_hints cl_khr_create_command_queue cl_intel_subgroups_char cl_intel_subgroups_long cl_khr_il_program cl_intel_mem_force_host_memory cl_khr_subgroup_extended_types cl_khr_subgroup_non_uniform_vote cl_khr_subgroup_ballot cl_khr_subgroup_non_uniform_arithmetic cl_khr_subgroup_shuffle cl_khr_subgroup_shuffle_relative cl_khr_subgroup_clustered_reduce cl_intel_device_attribute_query cl_khr_suggested_local_work_size cl_intel_split_work_group_barrier cl_khr_fp64 cl_khr_subgroups cl_intel_spirv_device_side_ cl_intel_spirv_media_block_io cl_intel_spirv_subgroups cl_khr_spirv_no_integer_wrap_decoration cl_intel_unified_shared_memory cl_khr_mipmap_image cl_khr_mipmap_image_writes cl_intel_planar_yuv cl_intel_packed_yuv cl_intel_motion_estimation cl_intel_device_side_avc_mo cl_intel_advanced_motion_estimation cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_image2d_from_buffer cl_khr_depth_images cl_khr_3d_image_writes cl_intel_media_block_io cl_intel_va_api_media_sharing cl_intel_sharing_format_query cl_khr_pci_bus_info Platform Numeric Version Platform Extensions function suffix Platform Host timer resolution Platform Name Number of devices Device Name Device Vendor Device Vendor ID Device Version Device UUID Driver UUID Valid Device LUID Device LUID Device Node Mask Device Numeric Version Driver Version Device OpenCL C Version Device OpenCL C all versions OpenCL C OpenCL C OpenCL C Device OpenCL C features __opencl_c_3d_image_writes __opencl_c_images __opencl_c_read_write_images __opencl_c_atomic_order_acq_rel __opencl_c_atomic_order_seq_cst __opencl_c_atomic_scope_all_devices __opencl_c_atomic_scope_device __opencl_c_generic_address_space __opencl_c_program_scope_gl __opencl_c_work_group_colle __opencl_c_subgroups __opencl_c_pipes __opencl_c_fp64 Latest comfornace test passed Device Type Device Profile Device Available Compiler Available Linker Available Max compute units Max clock frequency Device Partition Max number of sub-devices Supported partition types Supported affinity domains Max work item dimensions Max work item sizes Max work group size Preferred work group Preferred work group Max sub-groups per work group Sub-group sizes (Intel) Preferred / native vector sizes
char short int long half float double Half-precision Floating-point support Denormals Infinity and NANs Round to nearest Round to zero Round to infinity IEEE754-2008 fused multiply-add Support is emulated in software Single-precision Floating-point support Denormals Infinity and NANs Round to nearest Round to zero Round to infinity IEEE754-2008 fused multiply-add Support is emulated in software Correctly-rounded Double-precision Floating-point support Denormals Infinity and NANs Round to nearest Round to zero Round to infinity IEEE754-2008 fused multiply-add Support is emulated in software Address bits Global memory size Error Correction support Max memory allocation Unified memory for Host and Device Shared Virtual Memory Coarse-grained buffer sharing Fine-grained buffer sharing Fine-grained system sharing Atomics Minimum alignment for any data type Alignment of base address Preferred alignment for atomics
SVM Global Local Atomic memory capabilities Atomic fence capabilities Max size for global variable Preferred total size of global vars Global Memory cache type Global Memory cache size Global Memory cache line size Image support Max number of samplers per kernel Max size for 1D images from buffer Max 1D or 2D image array size Base address alignment Pitch alignment for 2D image buffers Max 2D image size Max planar YUV image size Max 3D image size Max number of read image args Max number of write image args Max number of read/write image args Pipe support Max number of pipe args Max active pipe reservations Max pipe packet size Local memory type Local memory size Max number of constant args Max constant buffer size Generic address space support Max size of kernel argument Queue properties (on host)
Out-of-order execution Profiling Device enqueue capabilities Queue properties (on device)
Out-of-order execution Profiling Preferred size Max size Max queues on device Max events on device Prefer user sync for interop Profiling timer resolution Execution capabilities
Run OpenCL kernels Run native kernels Non-uniform work-groups Work-group collective functions Sub-group independent forward progress IL version ILs with version SPIR versions printf() buffer size Built-in kernels Built-in kernels with version block_advanced_motion_estim block_advanced_motion_estim Motion Estimation Device-side AVC Motion Supports texture sampler use Supports preemption Device Extensions Device Extensions with Version cl_khr_device_uuid cl_khr_fp16 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_intel_command_queue_families cl_intel_subgroups cl_intel_required_subgroup_size cl_intel_subgroups_short cl_khr_spir cl_intel_accelerator cl_intel_driver_diagnostics cl_khr_priority_hints cl_khr_throttle_hints cl_khr_create_command_queue cl_intel_subgroups_char cl_intel_subgroups_long cl_khr_il_program cl_intel_mem_force_host_memory cl_khr_subgroup_extended_types cl_khr_subgroup_non_uniform_vote cl_khr_subgroup_ballot cl_khr_subgroup_non_uniform_arithmetic cl_khr_subgroup_shuffle cl_khr_subgroup_shuffle_relative cl_khr_subgroup_clustered_reduce cl_intel_device_attribute_query cl_khr_suggested_local_work_size cl_intel_split_work_group_barrier cl_khr_fp64 cl_khr_subgroups cl_intel_spirv_device_side_ cl_intel_spirv_media_block_io cl_intel_spirv_subgroups cl_khr_spirv_no_integer_wrap_decoration cl_intel_unified_shared_memory cl_khr_mipmap_image cl_khr_mipmap_image_writes cl_intel_planar_yuv cl_intel_packed_yuv cl_intel_motion_estimation cl_intel_device_side_avc_mo cl_intel_advanced_motion_estimation cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_image2d_from_buffer cl_khr_depth_images cl_khr_3d_image_writes cl_intel_media_block_io cl_intel_va_api_media_sharing cl_intel_sharing_format_query cl_khr_pci_bus_info NULL platform behavior
clGetPlatformInfo(NULL, clGetDeviceIDs(NULL, clCreateContext(NULL, …) [default] clCreateContextFromType(NULL, Platform Name Device Name clCreateContextFromType(NULL, clCreateContextFromType(NULL, Platform Name Device Name clCreateContextFromType(NULL, clCreateContextFromType(NULL, clCreateContextFromType(NULL, Platform Name Device Name ICD loader properties
ICD loader Name ICD loader Vendor ICD loader Version ICD loader Profile 1
Intel(R) OpenCL HD Graphics
Intel(R) Corporation
OpenCL 3.0
FULL_PROFILE
cl_khr_byte_addressable_store cl_khr_device_uuid cl_khr_fp16 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_intel_command_queue_families cl_intel_subgroups cl_intel_required_subgroup_size cl_intel_subgroups_short cl_khr_spir cl_intel_accelerator cl_intel_driver_diagnostics cl_khr_priority_hints cl_khr_throttle_hints cl_khr_create_command_queue cl_intel_subgroups_char cl_intel_subgroups_long cl_khr_il_program cl_intel_mem_force_host_memory cl_khr_subgroup_extended_types cl_khr_subgroup_non_uniform_vote cl_khr_subgroup_ballot cl_khr_subgroup_non_uniform_arithmetic cl_khr_subgroup_shuffle cl_khr_subgroup_shuffle_relative cl_khr_subgroup_clustered_reduce cl_intel_device_attribute_query cl_khr_suggested_local_work_size cl_intel_split_work_group_barrier cl_khr_fp64 cl_khr_subgroups cl_intel_spirv_device_side_avc_motion_estimation cl_intel_spirv_media_block_io cl_intel_spirv_subgroups cl_khr_spirv_no_integer_wrap_decoration cl_intel_unified_shared_memory cl_khr_mipmap_image cl_khr_mipmap_image_writes cl_intel_planar_yuv cl_intel_packed_yuv cl_intel_motion_estimation cl_intel_device_side_avc_motion_estimation cl_intel_advanced_motion_estimation cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_image2d_from_buffer cl_khr_depth_images cl_khr_3d_image_writes cl_intel_media_block_io cl_intel_va_api_media_sharing cl_intel_sharing_format_query cl_khr_pci_bus_info
cl_khr_byte_addressable_store 0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
avc_motion_estimation 0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
tion_estimation 0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0xc00000 (3.0.0)
INTEL
1ns
Intel(R) OpenCL HD Graphics
1
Intel(R) UHD Graphics 620 [0x5917]
Intel(R) Corporation
0x8086
OpenCL 3.0 NEO
86800000-1759-0000-0000-000000000000
32322e34-332e-3330-0000-000000000000
No
2017-825bfe7f0000
0
0xc00000 (3.0.0)
22.43.30
OpenCL C 1.2
OpenCL C 0x400000 (1.0.0)
0x401000 (1.1.0)
0x402000 (1.2.0)
0xc00000 (3.0.0)
__opencl_c_int64 0xc00000 (3.0.0)
0xc00000 (3.0.0)
0xc00000 (3.0.0)
0xc00000 (3.0.0)
0xc00000 (3.0.0)
0xc00000 (3.0.0)
0xc00000 (3.0.0)
0xc00000 (3.0.0)
0xc00000 (3.0.0)
obal_variables 0xc00000 (3.0.0)
ctive_functions 0xc00000 (3.0.0)
0xc00000 (3.0.0)
0xc00000 (3.0.0)
0xc00000 (3.0.0)
v2022-04-22-00
GPU
FULL_PROFILE
Yes
Yes
Yes
24
1100MHz
(core)
0
None
(n/a)
3
256x256x256
256
size multiple (device) 32
size multiple (kernel) 32
32
8, 16, 32
16 / 16
8 / 8
4 / 4
1 / 1
8 / 8 (cl_khr_fp16)
1 / 1
1 / 1 (cl_khr_fp64)
(cl_khr_fp16)
Yes
Yes
Yes
Yes
Yes
Yes
No
(core)
Yes
Yes
Yes
Yes
Yes
Yes
No
divide and sqrt operations Yes
(cl_khr_fp64)
Yes
Yes
Yes
Yes
Yes
Yes
No
64, Little-Endian
13319065600 (12.4GiB)
No
4294959104 (4GiB)
Yes
(SVM) capabilities (core)
Yes
No
No
No
128 bytes
1024 bits (128 bytes)
64 bytes
64 bytes
64 bytes
relaxed, acquire/release, sequentially-consistent, work-group scope, device scope, all-devices scope
relaxed, acquire/release, sequentially-consistent, work-item scope, work-group scope, device scope, all-devices scope
65536 (64KiB)
4294959104 (4GiB)
Read/Write
786432 (768KiB)
64 bytes
Yes
16
268434944 pixels
2048 images
for 2D image buffers 4 bytes
4 pixels
16384x16384 pixels
16384x16352 pixels
16384x16384x2048 pixels
128
128
128
Yes
16
1
1024
Local
65536 (64KiB)
8
4294959104 (4GiB)
Yes
2048 (2KiB)
Yes
Yes
(n/a)
No
No
0
0
0
0
Yes
83ns
Yes
No
Yes
Yes
Yes
SPIR-V_1.2
SPIR-V 0x402000 (1.2.0)
1.2
4194304 (4MiB)
block_motion_estimate_intel;block_advanced_motion_estimate_check_intel;block_advanced_motion_estimate_bidirectional_check_intel;
block_motion_estimate_intel 0x400000 (1.0.0)
ate_check_intel 0x400000 (1.0.0)
ate_bidirectional_check_intel 0x400000 (1.0.0)
accelerator version (Intel) 2
Estimation version 1
Yes
No
cl_khr_byte_addressable_store cl_khr_device_uuid cl_khr_fp16 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_intel_command_queue_families cl_intel_subgroups cl_intel_required_subgroup_size cl_intel_subgroups_short cl_khr_spir cl_intel_accelerator cl_intel_driver_diagnostics cl_khr_priority_hints cl_khr_throttle_hints cl_khr_create_command_queue cl_intel_subgroups_char cl_intel_subgroups_long cl_khr_il_program cl_intel_mem_force_host_memory cl_khr_subgroup_extended_types cl_khr_subgroup_non_uniform_vote cl_khr_subgroup_ballot cl_khr_subgroup_non_uniform_arithmetic cl_khr_subgroup_shuffle cl_khr_subgroup_shuffle_relative cl_khr_subgroup_clustered_reduce cl_intel_device_attribute_query cl_khr_suggested_local_work_size cl_intel_split_work_group_barrier cl_khr_fp64 cl_khr_subgroups cl_intel_spirv_device_side_avc_motion_estimation cl_intel_spirv_media_block_io cl_intel_spirv_subgroups cl_khr_spirv_no_integer_wrap_decoration cl_intel_unified_shared_memory cl_khr_mipmap_image cl_khr_mipmap_image_writes cl_intel_planar_yuv cl_intel_packed_yuv cl_intel_motion_estimation cl_intel_device_side_avc_motion_estimation cl_intel_advanced_motion_estimation cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_image2d_from_buffer cl_khr_depth_images cl_khr_3d_image_writes cl_intel_media_block_io cl_intel_va_api_media_sharing cl_intel_sharing_format_query cl_khr_pci_bus_info
cl_khr_byte_addressable_store 0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
avc_motion_estimation 0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
tion_estimation 0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
0x400000 (1.0.0)
CL_PLATFORM_NAME, …) Intel(R) OpenCL HD Graphics
CL_DEVICE_TYPE_ALL, …) Success [INTEL]
Success [INTEL]
CL_DEVICE_TYPE_DEFAULT) Success (1)
Intel(R) OpenCL HD Graphics
Intel(R) UHD Graphics 620 [0x5917]
CL_DEVICE_TYPE_CPU) No devices found in platform
CL_DEVICE_TYPE_GPU) Success (1)
Intel(R) OpenCL HD Graphics
Intel(R) UHD Graphics 620 [0x5917]
CL_DEVICE_TYPE_ACCELERATOR) No devices found in platform
CL_DEVICE_TYPE_CUSTOM) No devices found in platform
CL_DEVICE_TYPE_ALL) Success (1)
Intel(R) OpenCL HD Graphics
Intel(R) UHD Graphics 620 [0x5917]
OpenCL ICD Loader
OCL Icd free software
2.3.1
OpenCL 3.0

Francewhoa · April 8, 2024, 10:03pm

For more recent 2024 Linux, for faster calculation in LibreOffice I suggest considering those two options to choose from:

Option 1. Mesa Only.

Steps:

If not already done, install the appropriate Mesa packages for your Linux
- Most modern Linux already use Mesa by default. For example, on Debian 12 Bookworm, Mesa is both installed and activated by default. Easy.
- For example, for Debian 12 Bookworm:
  1. If not already done, install those two packages clinfo and mesa-opencl-icd
  2. Execute this command clinfo, without sudo permission, to test that Mesa for OpenCL is installed. In the long result, search for the two lines Device Name and Device Version.
Using LibreOffice, activate OpenCL
- If somehow OpenCL can’t be quickly and easily activated on your Linux, I suggest the alternative 2nd option below.

Option 2. LibreOffice 24 + Threaded.

Steps:

For potentially faster performance than OpenCL, simply leave OpenCL deactivated but install the latest LibreOffice version 24 or more recent using this official and safe repository at https://flathub.org/apps/org.libreoffice.LibreOffice
- This repository above has very good and stable support for the automated “Threaded” mode. Which means that the more core your CPU has, the faster the calculations are. For example, with a 16 core CPU, calculations are automatically up to 16 time faster. So Threaded is potentially faster then OpenCL. On most modern Linux, with LibreOffice 24, Threaded is activated by default. Easy
To verify that your LibreOffice Calc is Threaded, using Calc, navigate to Help menu —> About LibreOffice option. This screenshot shows 8 CPU cores. So this Calc complete calculation up to 8 times faster. In other words, potentially faster then OpenCL.

LO-24-Threaded-Screenshot-441086×699 149 KB

Below is the same suggestion as above. But with details for those interested.

What is Mesa?

Starting in 2024, to activate OpenCL for LibreOffice version 24 or more recent, I suggest using the fully free and Libre Source (Open Source) Mesa created by https://www.mesa3d.org

Most modern Linux already use Mesa by default. And you do not need to install any additional driver. For example, on Debian 12 Bookworm, Mesa is both installed and activated by default. Easy. If somehow OpenCL can’t be quickly and easily activated on your Linux, I suggest the alternative 2nd option below.

For example, if you use Debian 12 Bookworm, install the appropriate package(s) below:

For Radeon GPU
mesa-opencl-icd
For NVIDIA GPU
nvidia-opencl-icd
For Intel GPU
intel-opencl-icd

For those interested in stronger security and stronger privacy, I suggest considering alternative to Nvidia Graphics Drivers which both fully libre source (open source) and supported by Debian. Such as c. For Debian, the Mesa documentation is at https://wiki.debian.org/Mesa The 3 packages listed above are example of Mesa.

For those not familiar with non-Mesa Nvidia Graphics Drivers, they are partly secret source (closed source/proprietary) software, owned by a for profit corporation and not supported by Debian. In other words, the non-Mesa Nvidia Graphics Drivers have weaker security and weaker privacy.

What is Threaded?

If you’re not familiar with Threaded it means that all CPU cores will automatically be used for LibreOffice calculations. For example, if your CPU has 8 or 24 core, then with threaded automatically activated, the calculations will be up to 8 to 24 times faster In comparison, some older version of LibreOffice did not support “Threaded”. In turn, only 1 CPU core was use for all calculation. In other words, the more CPU core you have, the faster the calculation are. This is all automated.

If you’re not familiar with using the Flathub repository to install the latest LibreOffice 24 or more recent, the documentation to install Flathub it is at https://flathub.org/setup After this one time setup, you get free and frequent updates. So your LibreOffice is always updated to the latest version available. Faster and faster With automatically activated Threaded for way faster calculation with large documents.