EngineX-Ascend/enginex-ascend-910-llama.cpp
10
0
Fork 0
Code Issues Pull Requests Actions 4 Projects Releases Wiki Activity

ggml-backend : add GGML_BACKEND_DEVICE_TYPE_IGPU device type (#15797)

Browse Source 
* ggml-backend : add GGML_BACKEND_DEVICE_TYPE_IGPU device type

ggml-backend : add device id to device props

llama : only use iGPU devices if there are no GPU devices

llama : do not use multiple devices from different backends with the same device id
This commit is contained in:
Diego Devesa
2025-09-11 13:47:38 -07:00
committed by GitHub
parent 0e6ff0046f
commit 360d6533db
7 changed files with 81 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
ggml/src/ggml-vulkan/ggml-vulkan.cpp
View File

@@ -12113,6 +12113,7 @@ static ggml_backend_buffer_type_t ggml_backend_vk_device_get_host_buffer_type(gg
static enum ggml_backend_dev_type ggml_backend_vk_device_get_type(ggml_backend_dev_t dev) {
UNUSED(dev);
// TODO: return GGML_BACKEND_DEVICE_TYPE_IGPU for integrated GPUs
return GGML_BACKEND_DEVICE_TYPE_GPU;
}
@@ -12120,6 +12121,7 @@ static void ggml_backend_vk_device_get_props(ggml_backend_dev_t dev, struct ggml
props->name = ggml_backend_vk_device_get_name(dev);
props->description = ggml_backend_vk_device_get_description(dev);
props->type = ggml_backend_vk_device_get_type(dev);
// TODO: set props->device_id to PCI bus id
ggml_backend_vk_device_get_memory(dev, &props->memory_free, &props->memory_total);
props->caps = {
/* .async = */ false,