issues
search
nod-ai
/
iree-amd-aie
IREE plugin repository for the AMD AIE accelerator
Apache License 2.0
69
stars
30
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
write a script to build just XDNA kernel module
#869
makslevental
opened
2 weeks ago
0
Bump mlir-air to 3d1a4e19ff748897a37e0eb88b59999197dbc0f8
#868
erwei-xilinx
closed
2 weeks ago
0
Pass to align transfer_reads
#867
newling
closed
2 weeks ago
3
Towards vectorized convolution (second PR)
#866
newling
closed
3 weeks ago
0
MacOS 12 deprecated, bump to 13
#865
jtuyls
closed
2 weeks ago
0
Towards vectorized convolution (first PR)
#864
newling
closed
3 weeks ago
3
Bump IREE to 3cf5b65f736ce50c9890190b80e6343c0b929d56
#863
yzhang93
closed
3 weeks ago
0
[LinalgFunctionOutlining] Create a pass to outline linalg compute ops
#862
Abhishek-Varma
closed
2 weeks ago
0
Port `drm_timeline_syncobj` from xdna
#861
makslevental
opened
3 weeks ago
0
Bump driver dependency
#860
jtuyls
closed
3 weeks ago
0
[ Do not review] For test
#859
dezhiAmd
opened
3 weeks ago
1
[XRT-LITE] disable xrt-lite cts by default (but enable in CI)
#858
makslevental
closed
3 weeks ago
0
Bump mlir-air to 3d1a4e19ff748897a37e0eb88b59999197dbc0f8
#857
erwei-xilinx
closed
3 weeks ago
0
[Vectorization][ObjectFifo] Enable larger Matmul + Truncf
#856
Abhishek-Varma
closed
2 weeks ago
5
Bump mlir-air; more generic air compiler pipeline
#855
erwei-xilinx
closed
4 weeks ago
0
[Insert-Loops-For-Vec] Update insert-loops-for-vectorization pass
#854
Abhishek-Varma
closed
3 weeks ago
1
Bump mlir-air to 24cb14e6d2233e819a5455928e4237ef319e6fc8
#853
erwei-xilinx
closed
1 month ago
0
[HAL] remove unused headers from cts
#852
makslevental
closed
1 month ago
0
[XRT-LITE] add ability to configure NPU power mode
#851
makslevental
closed
1 month ago
0
[SplitLogicalObjectFifos] Add support for dma tranposed on the target side
#850
yzhang93
closed
3 weeks ago
0
Only output bo needs to be synced from device after result is available
#849
dezhiAmd
closed
1 month ago
8
Bump IREE to df5e5aab044ed5b6c5860b0b291c95eafe1c2522
#848
makslevental
closed
1 month ago
0
`iree-codegen-iree-comprehensive-bufferize` genereates `memref`s with dynamic offset
#847
makslevental
closed
1 month ago
7
Increase the K tile size in L1 for matmul ops
#846
yzhang93
closed
1 week ago
12
Bump IREE to 05bbcf1385146d075829cd940a52bf06961614d0
#845
makslevental
closed
1 month ago
2
[DO NOT REVIEW YET] [WIP] Enable vectorization after bufferization (insert-cores)
#844
Abhishek-Varma
closed
3 weeks ago
1
[POC][WIP] Merge prologue and epilogue into main loop
#843
newling
closed
3 days ago
0
[WIP] Bump XRT
#842
newling
closed
1 month ago
1
[AMDAIEDistributeCoresAndObjectFifos] Factorize out memory privatization sub-pass
#841
newling
closed
1 month ago
0
Add controlcode to transaction lowering.
#840
jtuyls
closed
1 month ago
1
[CombineStridedOps] Add a combinable case
#839
yzhang93
closed
1 month ago
0
Move most of LLVM lowering out of aie2xclbin
#838
newling
closed
1 month ago
0
Bump IREE 10/09/2024
#837
makslevental
closed
1 month ago
5
[AMDAIECoreToStandard] Pass simplification
#836
newling
closed
1 month ago
0
[Testing] Fix bf16->f32 utility to preserve shape of the input
#835
Abhishek-Varma
closed
1 month ago
1
Temporarily disable double buffering for better performance
#834
yzhang93
closed
1 month ago
2
Run aievec lowering passes in 'main' pipeline
#833
newling
closed
1 month ago
3
[Cleanup] Remove public API without implementation
#832
newling
closed
1 month ago
0
[AMDAIETemporaryAllocBufferization] Don't change operand of dealloc
#831
newling
closed
1 month ago
0
[Testing] e2e numerical test with user defined inputs and expected value
#830
newling
closed
1 month ago
0
[CI] Support bf16 output in end-to-end numerical testing
#829
newling
closed
1 month ago
8
Optimize double buffering and loop pipelining
#828
yzhang93
opened
1 month ago
0
[NpuDmaCpyNdOp/NpuDmaWaitOp] Return optional async token and wait for multiple
#827
jtuyls
closed
1 month ago
0
[DmaLoopSubsumption] Relax circular dma loop subsumption condition
#826
yzhang93
closed
1 month ago
0
Remove load alignment removal pass
#825
newling
closed
1 month ago
1
[AMDAIETemporaryAllocBufferization] Bufferize temporary allocs that aren't defined in core
#824
newling
closed
1 month ago
0
[Objectfifo] Create a new pass to flatten vectorized ops
#823
Abhishek-Varma
closed
1 month ago
2
[Matmul+Truncf] Enable Matmul+Truncf for shorter shape on Pack-Peel + Objectfifo
#822
Abhishek-Varma
closed
1 month ago
4
Multicore convolution
#821
newling
opened
1 month ago
2
Numerics issue with vectorized conv2d
#820
newling
closed
2 weeks ago
3
Previous
Next