Skip to content

SYCL: virtual-memory device pool off by default #24673

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
sanmai wants to merge 1 commit into from
+3 −3
Closed

SYCL: virtual-memory device pool off by default #24673

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion docs/backend/SYCL.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -715,7 +715,7 @@ 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. |
| GGML_SYCL_ENABLE_VMM | 0 (default) or 1 | 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. |

Expand Down
4 changes: 2 additions & 2 deletions ggml/src/ggml-sycl/ggml-sycl.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -77,7 +77,7 @@ int g_ggml_sycl_debug = 0;
int g_ggml_sycl_disable_optimize = 0;
int g_ggml_sycl_disable_graph = 0;
int g_ggml_sycl_disable_dnn = 0;
int g_ggml_sycl_enable_vmm = 1;
int g_ggml_sycl_enable_vmm = 0;
int g_ggml_sycl_prioritize_dmmv = 0;
int g_ggml_sycl_use_async_mem_op = 0;
int g_ggml_sycl_use_async_mem_op_requested = 1;
Expand Down Expand Up @@ -265,7 +265,7 @@ static void ggml_check_sycl() try {
g_ggml_sycl_disable_optimize = ggml_sycl_get_env("GGML_SYCL_DISABLE_OPT", 0);
g_ggml_sycl_disable_graph = ggml_sycl_get_env("GGML_SYCL_DISABLE_GRAPH", 1);
g_ggml_sycl_disable_dnn = ggml_sycl_get_env("GGML_SYCL_DISABLE_DNN", 0);
g_ggml_sycl_enable_vmm = ggml_sycl_get_env("GGML_SYCL_ENABLE_VMM", 1);
g_ggml_sycl_enable_vmm = ggml_sycl_get_env("GGML_SYCL_ENABLE_VMM", 0);
g_ggml_sycl_prioritize_dmmv = ggml_sycl_get_env("GGML_SYCL_PRIORITIZE_DMMV", 0);

#ifdef SYCL_FLASH_ATTN
Expand Down
Loading