Skip to content

opencl: Add MxFP4 MoE bin kernel for Adreno X1 devices - #30691

Merged
max-krasnyansky merged 4 commits into
ggml-org:masterfrom
qualcomm:sg/moe-gemm-mxfp4-ldck
Oct 16, 2026
Merged

max-krasnyansky merged 4 commits into
ggml-org:masterfrom
qualcomm:sg/moe-gemm-mxfp4-ldck

Conversation

@shawngu-quic

Copy link
Copy Markdown
Contributor

Overview

Add MxFP4 MoE bin kernel specifically optimized for Adreno X1 devices.

Additional information

Requirements

@shawngu-quic
shawngu-quic requested a review from a team as a code owner October 16, 2026 18:42
@github-actions github-actions Bot added ggml changes relating to the ggml tensor library for machine learning OpenCL Issues specific to the OpenCL backend labels Oct 16, 2026
@lhez

lhez commented Oct 16, 2026

Copy link
Copy Markdown
Contributor

@max-krasnyansky for a second review

@max-krasnyansky
max-krasnyansky merged commit 2f2763d into ggml-org:master Oct 16, 2026
31 of 33 checks passed
wanghqc added a commit to qualcomm/llama.cpp that referenced this pull request Oct 17, 2026
…30691)

Kernel library 0.0.15 adds gemm_moe_mxfp4_q8_1_dp4a_ila_ldck, served only
to X1 devices, which reads per-tile transposed q8_1 activations. Load it
when present, compile the transpose kernel only then, and route mxfp4 MoE
prefill to it ahead of the source dp4a GEMM. X1 already ran the source
dp4a GEMM by default here, so upstream's X1E default flip is not needed.

GGML_OPENCL_MOE_MXFP4_LDCK_BIN=0 keeps the source GEMM;
GGML_OPENCL_MOE_DISPATCH_LOG=1 now also names the mxfp4 route.
edwardyoon pushed a commit to edwardyoon/focus-llama that referenced this pull request Oct 24, 2026
* opencl: add bin kernel `gemm_moe_mxfp4_q8_1_dp4a_ila_ldck` for X1

* opencl: fix whitespace

* opencl: undo unnecessary change

* opencl: enable dp4a clc for X1E

---------

Co-authored-by: Li He <lih@qti.qualcomm.com>
auRose94 pushed a commit to auRose94/llama.cpp that referenced this pull request Oct 28, 2026
* opencl: add bin kernel `gemm_moe_mxfp4_q8_1_dp4a_ila_ldck` for X1

* opencl: fix whitespace

* opencl: undo unnecessary change

* opencl: enable dp4a clc for X1E

---------

Co-authored-by: Li He <lih@qti.qualcomm.com>
@BrewTestBot BrewTestBot mentioned this pull request Oct 28, 2026
1 task done
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

ggml changes relating to the ggml tensor library for machine learning OpenCL Issues specific to the OpenCL backend

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants