happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

cuda : fix UMA detection on discrete GPUs. (#17537)

This commit is contained in:
matt23654 2025-11-27 11:35:35 +00:00 committed by GitHub
parent cd8370b408
commit 909072abcf
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
ggml/src/ggml-cuda/ggml-cuda.cu
View File

@ -3837,7 +3837,7 @@ static void ggml_backend_cuda_device_get_memory(ggml_backend_dev_t dev, size_t *
// Check if UMA is explicitly enabled via environment variable // Check if UMA is explicitly enabled via environment variable
bool uma_env = getenv("GGML_CUDA_ENABLE_UNIFIED_MEMORY") != nullptr; bool uma_env = getenv("GGML_CUDA_ENABLE_UNIFIED_MEMORY") != nullptr;
bool is_uma = prop.unifiedAddressing > 0 || uma_env; bool is_uma = prop.integrated > 0 || uma_env;
if (is_uma) { if (is_uma) {
// For UMA systems (like DGX Spark), use system memory info // For UMA systems (like DGX Spark), use system memory info