geohot
9d0cd0ebcb
hotfix: switch benchmark to qwen3.8
2026-08-23 20:34:30 -07:00
YassineYousfi and GitHub
11edcc144c
disk cache: thread-local db conn ( #17694 )
2026-08-23 19:18:47 -07:00
chenyu and GitHub
a31aca9e52
don't hardcode dtype.int const in decode_hevc_frame ( #17696 )
2026-08-23 22:00:41 -04:00
chenyu and GitHub
8b164aefea
test update from weak const ( #17689 )
2026-08-23 17:30:16 -04:00
George Hotz and GitHub
477b573807
update llm kv cache to be half ( #17690 )
...
* update llm kv cache to be half / chunk_size to always be 32
* just dtype
2026-08-23 14:02:54 -07:00
geohot
bb0e99acbf
hotfix: skip that nan test on mac
2026-08-23 08:40:46 -07:00
nimlgen and GitHub
93865e2c66
hcq2: sunday housekeeping ( #17686 )
...
* hcq2: sunday housekeeping
* x
2026-08-23 17:20:40 +03:00
George Hotz and GitHub
5b60a09ab0
some fixes for the AMD emulator ( #17684 )
...
* some fixes for the AMD emulator
* simpler
* revert
* min
2026-08-22 22:48:07 -07:00
chenyu and GitHub
b0a1285330
lil decomp cleanup [PR] ( #17682 )
2026-08-22 18:11:14 -04:00
nimlgen and GitHub
a2e64e16aa
hcq2: early usb ( #17683 )
...
* hcq2: usb interface and submit
* x
* x
* x
* x
* r
* x
2026-08-23 00:22:39 +03:00
chenyu and GitHub
a9069c177a
make decomp pass SPEC=2 [PR] ( #17681 )
2026-08-22 08:01:15 -04:00
chenyu and GitHub
8950942e75
remove explicit dtype for NOOP and decomp [PR] ( #17678 )
2026-08-21 22:27:36 -04:00
chenyu and GitHub
356f665377
test update for weak const ( #17675 )
2026-08-21 21:50:29 -04:00
chenyu and GitHub
7204d46786
delete dtype_from_uop INDEX exempt ( #17674 )
2026-08-21 21:08:06 -04:00
George Hotz and GitHub
af242819d8
refactor the AMD emulator slop (kimi) ( #17673 )
...
* refactor the AMD emulator slop (kimi)
* mypy
2026-08-21 18:00:58 -07:00
wozeparrot and GitHub
52596dbf38
gptoss: fused ce ( #17672 )
2026-08-21 16:19:28 -07:00
sirhcm and GitHub
07cce78cec
compile3: log printed timings ( #17671 )
2026-08-21 19:18:32 -04:00
George Hotz and GitHub
f986829461
keep IndexingContext scoped in indexing ( #17670 )
2026-08-21 14:12:39 -07:00
nimlgen and GitHub
4fd4eafb23
nv: hevc ( #17661 )
...
* nv: hevc
* x
* nv: zero the nvdec scratch buffers
2026-08-21 23:37:09 +03:00
chenyu and GitHub
daa154aa22
FLOORDIV to SHR for powers of 2 [pr] ( #17669 )
2026-08-21 16:28:04 -04:00
nimlgen and GitHub
12f889aaad
hcq2: parallel compile ( #17667 )
...
* hcq2: parallel compile
* Dx
2026-08-21 23:27:21 +03:00
sirhcm and GitHub
298748ebd3
ci: remove setup-python ( #17665 )
2026-08-21 16:19:47 -04:00
George Hotz and GitHub
3082956a17
usb copyin: async arm and drain, 323 MB/s on comma ( #17663 )
...
* usb copyin: can safely arm before drain
* perf counter
* 294 MB/s for comma
* free speed with async transfers
2026-08-21 13:12:03 -07:00
chenyu and GitHub
402bea7ddd
l2i and sign_extend cleanups [pr] ( #17668 )
...
towards good threefry decomp
2026-08-21 15:55:45 -04:00
chenyu and GitHub
8f9cbdf0cc
few more self folding [pr] ( #17657 )
2026-08-21 15:21:28 -04:00
chenyu and GitHub
8f59041ee5
more wgsl pack cleanups [PR] ( #17664 )
2026-08-21 14:34:25 -04:00
nimlgen and GitHub
1cf8a2c7fe
hcq2: use shrink.bitcast ( #17653 )
...
* hcq2: shrink.bitcast
* x
* x
* x
* s
* x
* Dx
* Revert "hotfix: disable HCQ2"
This reverts commit a57188ea6d
2026-08-21 21:24:34 +03:00
chenyu and GitHub
3919ce8427
ceildiv in wgsl _packed_size [pr] ( #17662 )
2026-08-21 14:02:20 -04:00
756e82e055
usb amd: pipelined copyin over the 0xF2 engine (2.6x faster) ( #17628 )
...
Stream 240KB chunks into two alternating 256KB SRAM bounce windows; each chunk
ends in a unique 512B sentinel that a prebuilt SDMA ring polls before copying
the chunk to VRAM, followed by an in-order drain fence that the host waits on
before re-arming a window. No timing assumptions in either direction: the
sentinel is in-stream proof of data landing, the fence proves a full drain.
Adds a small pooled async bulk-OUT layer to USB3 so staging the next chunk
overlaps the wire, and a slot_start parameter to scsi_write for the second
window. 107 -> 276 MB/s copyin on tinyc8 (Kryo-3XX host).
Co-authored-by: tiny <tiny@local>
2026-08-21 10:38:50 -07:00
chenyu and GitHub
cc32aa18db
don't match casted const in const_folding_pat [pr] ( #17658 )
2026-08-21 12:16:07 -04:00
qazal and GitHub
77f698e55b
cleanup the mxfp4 gemm ( #17660 )
2026-08-22 00:59:04 +09:00
chenyu and GitHub
554d078ac4
fix render_marg [pr] ( #17656 )
2026-08-21 10:03:22 -04:00
chenyu and GitHub
176377ff6e
weak 1 for FDIV in get_late_rewrite_patterns [PR] ( #17655 )
2026-08-21 09:33:43 -04:00
qazal and GitHub
1c3c9e96f6
remove setUp from test_sqtt_profiler ( #17652 )
...
* remove setUp from test_sqtt_profiler
* remove that
* cleaner
* do not need that
2026-08-21 17:29:11 +09:00
qazal and GitHub
e8a8d99b99
profiler tracing using unique keys ( #17651 )
...
* profiler tracing using unique keys
* cleaner + comment
2026-08-21 17:06:01 +09:00
qazal and GitHub
dcc2d021e7
prereq viz cleanups for unique profile keys ( #17649 )
...
* cleaner
* just use VIZ=-2
* better
2026-08-21 16:18:25 +09:00
qazal and GitHub
80bf60d782
profiler failing test for non unique kernel names ( #17647 )
...
* err
* expected
2026-08-21 13:52:35 +09:00
George Hotz and GitHub
1cb0600086
fix llm vars regression (kimi) ( #17644 )
...
* fix llm regression (kimi)
* unneeded
* more principled
2026-08-20 21:31:43 -07:00
chenyu and GitHub
1bcb6bdc62
no-op weak.py reorder [PR] ( #17646 )
2026-08-20 23:56:06 -04:00
qazal and GitHub
d716d0d927
hotfix: qkv bw kernel requires hipcc and cdna4 ( #17645 )
2026-08-21 12:38:45 +09:00
b1tg and GitHub
9216aa494c
llm prefill failing test ( #17630 )
2026-08-20 18:58:34 -07:00
George Hotz and GitHub
9aa9e11301
compile kernels in parallel ( #17629 )
...
* compile kernels in parallel (slop)
* cleanups
* fixes
* hand cleanups
* no PARALLEL with VIZ
* ugh, we need to remove the n from names
* no hcq2 in that test
* main for abstractions3
* fix compile server to be fork safe
* fix num cpu threads in context
* just use a lock
* fix __main__ on spawn
* compileonce
* python3 is double
* xdist sizing
* parallel compile
* fetch supports extract
* fix issues
* revert compiler server to master, drop compileonce
* normal workflow
* PARALLEL=0 for QCOMCL compiletest
2026-08-20 16:41:45 -07:00
George Hotz and GitHub
3fdbb82bfe
add ansipad and PARALLEL contextvar ( #17642 )
2026-08-20 16:26:02 -07:00
George Hotz and GitHub
0ccef542e0
add extract support to fetch ( #17641 )
2026-08-20 16:20:16 -07:00
sirhcm and GitHub
c655aaf3a2
ci: venv in /opt/venv ( #17632 )
...
default python is now 3.14, not 3.12
2026-08-20 18:54:54 -04:00
chenyu and GitHub
592e3f8363
update const selector where folding [pr] ( #17640 )
...
folding a strong dtype WHERE to a weak const branch keeps the strong dtype
2026-08-20 18:49:34 -04:00
sirhcm and GitHub
3715006a21
fix float_to_bf16 on non-float32 inputs ( #17638 )
2026-08-20 18:14:30 -04:00
wozeparrot and GitHub
d80254c1d9
fa: remove swa recompute ( #17635 )
2026-08-20 14:26:44 -07:00
sirhcm and GitHub
c773891e3f
skip INT_MIN % -1 on X86 and LLVM ( #17637 )
2026-08-20 17:26:26 -04:00
chenyu and GitHub
0e7ab863a0
x86 REX issue ( #17580 )
...
Autogen / In-tree Autogen (comgr 2) (push) Successful in 54s
Unit Tests / Test LLM (push) Failing after 1m59s
Unit Tests / Fuzzing (push) Successful in 3m38s
Unit Tests / Docs (push) Successful in 3m51s
Deploy Docs / deploy (push) Successful in 3m59s
Unit Tests / hcq2 (push) Failing after 3m57s
Unit Tests / Linters (push) Successful in 4m4s
Unit Tests / Python Backend (push) Successful in 4m9s
Unit Tests / Models (push) Successful in 4m34s
Unit Tests / Linux (DSP) (push) Successful in 4m52s
Unit Tests / AMD ASM IDE (push) Successful in 5m23s
Unit Tests / Linux (DEV=CPU:X86) (push) Successful in 5m31s
Unit Tests / CL IMAGE Tests (push) Successful in 5m34s
Unit Tests / Linux (DEV=CL) (push) Successful in 5m47s
Unit Tests / Linux (DEV=WEBGPU) (push) Successful in 5m53s
Unit Tests / Compile-only (DEV=NULL:NAK:sm_120) (push) Successful in 1m54s
Unit Tests / Linux (DEV=CPU:LVP) (push) Successful in 6m0s
Unit Tests / SPEC=2 (2) (push) Successful in 6m7s
Unit Tests / Torch Backend Tests (push) Successful in 6m8s
Unit Tests / Torch Backend Training (push) Successful in 6m11s
Unit Tests / Linux (DEV=CPU:CLANG) (push) Successful in 6m13s
Unit Tests / Linux (amdllvm gfx1100) (push) Successful in 6m16s
Unit Tests / Linux (DEV=CPU:LLVM) (push) Successful in 6m18s
Unit Tests / openpilot Compile Tests (push) Successful in 6m25s
Unit Tests / Compile-only (DEV=NULL:IR3:a630) (push) Successful in 2m25s
Unit Tests / SPEC=2 (1) (push) Successful in 6m25s
Unit Tests / Unit Tests (push) Successful in 6m28s
Unit Tests / Null Tests (push) Successful in 6m31s
Unit Tests / Linux (amdllvm gfx1201) (push) Successful in 5m43s
Unit Tests / Linux (amd gfx1201) (push) Successful in 6m51s
Unit Tests / Linux (amd gfx1100) (push) Successful in 6m54s
Unit Tests / Linux (am) (push) Successful in 7m11s
Unit Tests / Linux (amdllvm gfx950) (push) Successful in 3m43s
Unit Tests / Optimization Tests (push) Successful in 7m26s
Unit Tests / Linux (ptx) (push) Successful in 3m35s
Unit Tests / Linux (amd gfx950) (push) Successful in 5m40s
Unit Tests / Linux (nv) (push) Successful in 4m4s
Unit Tests / ONNX (CPU) Tests (push) Successful in 9m34s
Autogen / In-tree Autogen (push) Successful in 10m55s
Unit Tests / Compile-only (DEV=NULL:QCOMCL:a630) (push) Failing after 8m29s
Autogen / In-tree Autogen (macos) (push) Canceled after 0s
Benchmarks / Mac pytest (push) Canceled after 0s
Benchmarks / LLM (DEV=AMD) (push) Canceled after 0s
Benchmarks / LLM (DEV=METAL) (push) Canceled after 0s
Benchmarks / LLM (DEV=NV) (push) Canceled after 0s
Benchmarks / HLB-CIFAR10 (DEV=AMD) (push) Canceled after 0s
Benchmarks / HLB-CIFAR10 (DEV=METAL) (push) Canceled after 0s
Benchmarks / HLB-CIFAR10 (DEV=NV) (push) Canceled after 0s
Benchmarks / MLPerf (AMD) (push) Canceled after 0s
Benchmarks / MLPerf (NV) (push) Canceled after 0s
Benchmarks / Stable Diffusion (DEV=AMD) (push) Canceled after 0s
Benchmarks / Stable Diffusion (DEV=METAL) (push) Canceled after 0s
Benchmarks / Stable Diffusion (DEV=NV) (push) Canceled after 0s
Benchmarks / Multi-GPU Benchmarks (DEV=AMD) (push) Canceled after 0s
Benchmarks / Multi-GPU Benchmarks (DEV=NV) (push) Canceled after 0s
Benchmarks / Tests (DEV=AMD) (push) Canceled after 0s
Benchmarks / Tests (DEV=METAL) (push) Canceled after 0s
Benchmarks / Tests (DEV=NV) (push) Canceled after 0s
Benchmarks / UsbGPU Benchmark (push) Canceled after 0s
Benchmarks / comma Benchmark (0.11.2) (push) Canceled after 0s
Benchmarks / comma Benchmark (0.11.0) (push) Canceled after 0s
Benchmarks / DSP Benchmark (push) Canceled after 0s
Benchmarks / UsbGPU Benchmark (comma) (push) Canceled after 0s
Benchmarks / PCI Driver Benchmark (DEV=AMD) (push) Canceled after 0s
Benchmarks / PCI Driver Benchmark (DEV=NV) (push) Canceled after 0s
Benchmarks / LLVM Speed (push) Canceled after 0s
Platform Tests / MacOS (unit) (push) Canceled after 0s
Platform Tests / MacOS (unit, mock) (push) Canceled after 0s
Platform Tests / MacOS (DEV=METAL) (1) (push) Canceled after 0s
Platform Tests / MacOS (DEV=METAL) (2) (push) Canceled after 0s
Platform Tests / MacOS (DEV=CPU:CLANG) (push) Canceled after 0s
Platform Tests / MacOS (DEV=CPU:LLVM) (push) Canceled after 0s
Platform Tests / MacOS (DEV=CPU:LVP) (push) Canceled after 0s
Platform Tests / MacOS (DEV=WEBGPU) (push) Canceled after 0s
Platform Tests / Windows (DEV=CPU:CLANG) (push) Canceled after 0s
Platform Tests / Windows (DEV=CPU:LLVM) (push) Canceled after 0s
Platform Tests / Windows (DEV=CPU:X86) (push) Canceled after 0s
Platform Tests / Windows (DEV=WEBGPU) (push) Canceled after 0s
* failing test
* fix
2026-08-20 16:49:04 -04:00
geohot
sirhcmandGitHub