nikola/llama_cpp
1
0
Fork 0
Code Issues Pull Requests Actions 18 Packages Projects Releases Wiki Activity

SYCL: implement ggml_sycl_pool_vmm (#22862)

Browse Source 
* SYCL: implement ggml_sycl_pool_vmm

* Add an option to bypass VMM with GGML_SYCL_DISABLE_VMM

* Clean up debugging logging

* document GGML_SYCL_DISABLE_VMM

* Multi-stream MoE optimization

* Revert "Multi-stream MoE optimization"

This reverts commit 938929c3f13a562ec67c59e87cc5d38595444cce.

* Update common.hpp

Co-authored-by: Neo Zhang <zhang.jianyu@outlook.com>

* Flip GGML_SYCL_DISABLE_VMM to GGML_SYCL_ENABLE_VMM

* add logging for GGML_SYCL_ENABLE_VMM when extension is not available (SYCL_EXT_ONEAPI_VIRTUAL_MEM macro)

* Apply suggestions from code review

Co-authored-by: Alexey Kopytko <alexey@kopytko.com>

* Apply suggestion from @sanmai

* Apply suggestion from @sanmai

---------

Co-authored-by: Neo Zhang <zhang.jianyu@outlook.com>
This commit is contained in:
Alexey Kopytko
2026-05-26 07:59:00 +03:00
committed by GitHub
co-authored by Neo Zhang
parent 7623de11d9
commit 581d020b12
3 changed files with 165 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/backend/SYCL.md
+2
View File
@@ -743,6 +743,7 @@ use 1 SYCL GPUs: [0] with Max compute units:512
| GGML_SYCL_DISABLE_GRAPH | 0 or 1 (default) | Disable running computations through SYCL Graphs feature. Disabled by default because SYCL Graph is still on development, no better performance. |
| GGML_SYCL_ENABLE_LEVEL_ZERO | 1 (default) or 0 | Use Level Zero API for device memory allocation instead of SYCL. Reduces system RAM usage on Intel dGPUs by avoiding DMA-buf/TTM host memory staging. Requires GGML_SYCL_SUPPORT_LEVEL_ZERO=ON at build time. |
| GGML_SYCL_DISABLE_DNN | 0 (default) or 1 | Disable running computations through oneDNN and always use oneMKL. |
| GGML_SYCL_ENABLE_VMM | 0 or 1 (default) | Enable the virtual-memory device pool. |
| ZES_ENABLE_SYSMAN | 0 (default) or 1 | Support to get free memory of GPU by sycl::aspect::ext_intel_free_memory.<br>Recommended to use when --split-mode = layer |
| UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS | 0 (default) or 1 | Allow SYCL/Unified Runtime Level Zero device allocations larger than 4 GiB. llama.cpp's direct Level Zero allocation path requests the relaxed maximum-size limit itself when GGML_SYCL_ENABLE_LEVEL_ZERO=1. |
@@ -753,6 +754,7 @@ Pass these via `CXXFLAGS` or add a one-off `#define` to enable a flag on the spo
| Name | Function |
|-----------------|----------------------------------------------------------------------------------|
| DEBUG_SYCL_POOL | Enable device memory pool logging on teardown. Useful for profiling allocations. |
| DEBUG_SYCL_MALLOC | Enable verbose per-call logging of device pool alloc/free operations. |
## Design Rule