Skip to content

Activity

Showing most recent first

Updated oneTBB version in driver install instructions

ProjectPhysXpushed 1 commit to master • a01b20e…14b670c • 
6 days ago

Updated Readme

ProjectPhysXpushed 1 commit to master • 970b8af…a01b20e • 
21 days ago

Fixed dual CU and IPC reporting on AMD RDNA1-4 GPUs, updated AMD driv…

ProjectPhysXpushed 1 commit to master • 03d08d3…970b8af • 
23 days ago

Fixed compiler warning with min_int

ProjectPhysXpushed 1 commit to master • 824d8ae…03d08d3 • 
on May 17

Disabled native dp4a in Intel CPU Runtime for OpenCL because it is sl…

ProjectPhysXpushed 1 commit to master • b0ba4d7…824d8ae • 
on May 17

Disabled native dp4a in Intel CPU Runtime for OpenCL because it is sl…

ProjectPhysXpushed 1 commit to master • 6355d58…b0ba4d7 • 
on May 17

Updated OpenCL driver installation guide and links

ProjectPhysXpushed 2 commits to master • 4d53e13…6355d58 • 
on Apr 18

Fixed bug in split_regex()

ProjectPhysXpushed 1 commit to master • 374a0c3…4d53e13 • 
on Mar 20

More robust dp4a detection

ProjectPhysXpushed 1 commit to master • be3b7f3…374a0c3 • 
on Mar 17

Fixed missing <chrono> header on some compilers

ProjectPhysXpushed 1 commit to master • 7f95637…be3b7f3 • 
on Mar 11

Also allow to use #undef in OpenCL C code

ProjectPhysXpushed 1 commit to master • c16d7ee…7f95637 • 
on Mar 1

Fixed broken CL_DEVICE_OPENCL_C_ALL_VERSIONS on AMD GPUs

ProjectPhysXpushed 1 commit to master • 2f3faca…c16d7ee • 
on Mar 1

OpenCL kernels are now compiled for latest supported OpenCL C standar…

ProjectPhysXpushed 1 commit to master • 3b18010…2f3faca • 
on Mar 1

Minor cosmetics

ProjectPhysXpushed 1 commit to master • c31c9ea…3b18010 • 
on Feb 26

Minor fix in print_error()

ProjectPhysXpushed 1 commit to master • 4509c66…c31c9ea • 
on Feb 26

Fixed compiling on macOS with new OpenCL headers

ProjectPhysXpushed 1 commit to master • 4105c61…4509c66 • 
on Feb 22

Renamed def_workgroup_size to cl_workgroup_size

ProjectPhysXpushed 1 commit to master • f061c99…4105c61 • 
on Feb 11

Better detection for nvidia__64_cores_per_cu

ProjectPhysXpushed 1 commit to master • 2836dd2…f061c99 • 
on Feb 9

Updated OpenCL headers, better device detection using vendor ID and N…

ProjectPhysXpushed 1 commit to master • da76bdb…2836dd2 • 
on Feb 8

Made make.sh executable

ProjectPhysXpushed 1 commit to master • 9084224…da76bdb • 
on Feb 5

Faster enqueueReadBuffer on modern CPUs with 64-Byte-aligned host_buf…

ProjectPhysXpushed 1 commit to master • 004b45b…9084224 • 
on Dec 27, 2024

Better VRAM capacity reporting correction for Intel dGPUs

ProjectPhysXpushed 1 commit to master • c7410be…004b45b • 
on Dec 24, 2024

Fixed TFlops estimate for Intel Battlemage GPUs

ProjectPhysXpushed 1 commit to master • 28b09a7…c7410be • 
on Dec 5, 2024

Fixed broken make.sh compile script

ProjectPhysXpushed 1 commit to master • 2f2453f…28b09a7 • 
on Nov 25, 2024

Automatically use zero-copy buffers on CPUs/iGPUs to reduce memory fo…

ProjectPhysXpushed 1 commit to master • bc440ca…2f2453f • 
on Nov 16, 2024

Minor cosmetics

ProjectPhysXpushed 1 commit to master • 789df22…bc440ca • 
on Sep 18, 2024

Enabled basic FP16 vector arithmetic support on Nvidia Pascal and new…

ProjectPhysXpushed 1 commit to master • 6392f3b…789df22 • 
on Aug 18, 2024

Fixed typo in Readme

ProjectPhysXpushed 1 commit to master • 38770d8…6392f3b • 
on Aug 7, 2024

Fixed maximum buffer allocation size limit for AMD GPUs

ProjectPhysXpushed 1 commit to master • d710c05…38770d8 • 
on Aug 3, 2024

Fixed maximum buffer allocation size limit in Intel CPU Runtime for O…

ProjectPhysXpushed 1 commit to master • 174b5a9…d710c05 • 
on Aug 3, 2024