[SYCL][Fusion] Kernel Fusion support for CUDA backend#8747

Merged

aelovikov-intel merged 25 commits intointel:syclfrom

sommerlukas:experiments/cuda-fusion

May 8, 2023

Commits on May 4, 2023

[SYCL][Fusion] Embed LLVM IR for SYCL for Nvidia
sommerlukas
committed
Enable LLVM IR as alternative fusion input format;
sommerlukas
committed
[SYCL][Fusion] Add result translation to PTX
sommerlukas
committed
[SYCL][Fusion] Provide correct target spec;
sommerlukas
committed
[SYCL][Fusion] Avoid removing dependencies
sommerlukas
committed
[SYCL][Fusion] Set device binary image format
sommerlukas
committed
[SYCL][Fusion] Refactor target-specific processing
sommerlukas
committed
[SYCL][Fusion] Do not require null terminator
sommerlukas
committed
[SYCL][Fusion] Refactor more target-specific code
sommerlukas
committed
[SYCL][Fusion] Handle attributes for CUDA fusion
sommerlukas
committed
[SYCL][Fusion] Cache and groom input binaries
sommerlukas
committed
[SYCL][Fusion] Disable heterogeneous ND ranges on CUDA
sommerlukas
committed
[SYCL][Fusion] Enable JIT caching for CUDA fusion
sommerlukas
committed
[SYCL][Fusion] Catch empty standard arguments
sommerlukas
committed
[SYCL][Fusion] Rebase and address feedback
sommerlukas
committed
[SYCL][Fusion] Update linkage graph diagram
sommerlukas
committed
Don't compile NVPTX-specifics if not supported
sommerlukas
committed
Migrate test changes from intel/llvm-test-suite
sommerlukas
committed
Address more PR feedback
sommerlukas
committed
Add test for kernel fusion with math function
sommerlukas
committed
Document CUDA kernel fusion in design documentation
sommerlukas
committed
Update kernel fusion design document
sommerlukas
committed
Fix formatting for test
sommerlukas
committed
Rebase on branch 'sycl'
sommerlukas
committed
Address PR feedback and formatting
sommerlukas
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[SYCL][Fusion] Kernel Fusion support for CUDA backend#8747

[SYCL][Fusion] Kernel Fusion support for CUDA backend#8747
aelovikov-intel merged 25 commits intointel:syclfrom
sommerlukas:experiments/cuda-fusion

Commits on May 4, 2023

[SYCL][Fusion] Embed LLVM IR for SYCL for Nvidia

Enable LLVM IR as alternative fusion input format;

[SYCL][Fusion] Add result translation to PTX

[SYCL][Fusion] Provide correct target spec;

[SYCL][Fusion] Avoid removing dependencies

[SYCL][Fusion] Set device binary image format

[SYCL][Fusion] Refactor target-specific processing

[SYCL][Fusion] Do not require null terminator

[SYCL][Fusion] Refactor more target-specific code

[SYCL][Fusion] Handle attributes for CUDA fusion

[SYCL][Fusion] Cache and groom input binaries

[SYCL][Fusion] Disable heterogeneous ND ranges on CUDA

[SYCL][Fusion] Enable JIT caching for CUDA fusion

[SYCL][Fusion] Catch empty standard arguments

[SYCL][Fusion] Rebase and address feedback

[SYCL][Fusion] Update linkage graph diagram

Don't compile NVPTX-specifics if not supported

Migrate test changes from intel/llvm-test-suite

Address more PR feedback

Add test for kernel fusion with math function

Document CUDA kernel fusion in design documentation

Update kernel fusion design document

Fix formatting for test

Rebase on branch 'sycl'

Address PR feedback and formatting

[SYCL][Fusion] Kernel Fusion support for CUDA backend#8747aelovikov-intel merged 25 commits intointel:syclintel/llvm:syclfrom sommerlukas:experiments/cuda-fusionsommerlukas/llvm:experiments/cuda-fusionCopy head branch name to clipboard

Commits

Commits on May 4, 2023

[SYCL][Fusion] Kernel Fusion support for CUDA backend#8747
aelovikov-intel merged 25 commits intointel:syclfrom
sommerlukas:experiments/cuda-fusion