Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

hexagon: add HTP kernel for GGML_OP_GATED_DELTA_NET (#22837)

Browse Source 
Implement the Gated Delta Net recurrence on HVX with:
- 4-row fused kernels for PP (prompt processing) path
- 8-row fused kernels for TG (token generation) path, reducing
  K/Q/gate vector reload overhead by 2x
- Separate PP/TG thread functions for I-cache isolation
- VTCM state scratchpad with DMA in/out for TG single-cycle access
- Vectorized gate exp via hvx_exp_f32
This commit is contained in:
Yanzhao Wang
2026-05-08 17:12:04 -07:00
committed by GitHub
parent c5703e03a5
commit 66001722aa
6 changed files with 1045 additions and 27 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-hexagon/htp/htp-ctx.h
+1
View File
@@ -106,5 +106,6 @@ int op_cumsum(struct htp_ops_context * octx);
int op_fill(struct htp_ops_context * octx);
int op_diag(struct htp_ops_context * octx);
int op_solve_tri(struct htp_ops_context * octx);
int op_gated_delta_net(struct htp_ops_context * octx);
#endif /* HTP_CTX_H */