metal : initial Metal4 tensor API support (#16634)

* metal : rework mat-mat multiplication

* metal : initial Metal4 support

* cont

* metal : detect tensor support

* cont : better ifdefs

* metal : support tensors in mul_mm_id

* metal : add env for disabling tensor API

* tests : restore

* metal : remove unused constants

* metal : fix check for bfloat tensor support

* cont : handle API incompatibilities

* cont : handle even more incompatibilities

* metal : use tensor API only on M5 and later
This commit is contained in:
Georgi Gerganov
2025-11-06 14:45:10 +02:00
committed by GitHub
parent b7f9010d24
commit 5b180c3d60
4 changed files with 617 additions and 147 deletions
+1 -4
View File
@@ -35,7 +35,6 @@ struct ggml_metal {
// additional, inference-time compiled pipelines
ggml_metal_pipelines_t pipelines_ext;
bool use_bfloat;
bool use_fusion;
bool use_concurrency;
bool use_graph_optimize;
@@ -121,11 +120,10 @@ ggml_metal_t ggml_metal_init(ggml_metal_device_t dev) {
}
}
const struct ggml_metal_device_props * props_dev = ggml_metal_device_get_props(dev);
//const struct ggml_metal_device_props * props_dev = ggml_metal_device_get_props(dev);
res->d_queue = dispatch_queue_create("ggml-metal", DISPATCH_QUEUE_CONCURRENT);
res->use_bfloat = props_dev->has_bfloat;
res->use_fusion = getenv("GGML_METAL_FUSION_DISABLE") == nil;
res->use_concurrency = getenv("GGML_METAL_CONCURRENCY_DISABLE") == nil;
@@ -147,7 +145,6 @@ ggml_metal_t ggml_metal_init(ggml_metal_device_t dev) {
memset(res->fuse_cnt, 0, sizeof(res->fuse_cnt));
GGML_LOG_INFO("%s: use bfloat = %s\n", __func__, res->use_bfloat ? "true" : "false");
GGML_LOG_INFO("%s: use fusion = %s\n", __func__, res->use_fusion ? "true" : "false");
GGML_LOG_INFO("%s: use concurrency = %s\n", __func__, res->use_concurrency ? "true" : "false");
GGML_LOG_INFO("%s: use graph optimize = %s\n", __func__, res->use_graph_optimize ? "true" : "false");