Skip to content

Add hacky M2 Max support #17

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

Open
initialed85 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Add hacky M2 Max support #17

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
3 changes: 2 additions & 1 deletion runtime/metal/DeviceCapabilities.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,8 @@ std::optional<std::string> DeviceCapabilities::validationError() const {
return "recommended_working_set_exceeds_physical_memory";
}
if (!maxBufferLengthBytes) return "max_buffer_length_unavailable";
if (appleGpuFamily < kMinimumAppleGpuFamily) return "apple_gpu_family_9_required";
// HACK(initialed85): attempting to add support for M2 Max
if (appleGpuFamily < kMinimumAppleGpuFamily) return "apple_gpu_family_8_required";
if (maxThreadgroupMemoryBytes < 32 * 1024) {
return "threadgroup_memory_below_32_kib";
}
Expand Down
3 changes: 2 additions & 1 deletion runtime/metal/DeviceCapabilities.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,8 @@ struct DeviceCapabilities {
uint32_t macosMinor = 0;
uint32_t macosPatch = 0;
// Highest supported MTLGPUFamilyAppleN.
static constexpr uint32_t kMinimumAppleGpuFamily = 9;
// HACK(initialed85): attempting to add support for M2 Max
static constexpr uint32_t kMinimumAppleGpuFamily = 8;
uint32_t appleGpuFamily = 0;
// IORegistry gpu-core-count; zero means unavailable. Kernel policy then
// assumes a large GPU when choosing parallelism.
Expand Down
3 changes: 2 additions & 1 deletion runtime/model/Runtime.mm
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -42,7 +42,8 @@
// Fixed reserves the memory plan carries beside the planned arenas: Metal
// pipeline objects and encoder scratch, and the process's own runtime overhead.
constexpr uint64_t kPipelineReserveBytes = 256ULL << 20;
constexpr uint64_t kRuntimeOverheadReserveBytes = 512ULL << 20;
// HACK(initialed85): trying to add support for M2 Max
constexpr uint64_t kRuntimeOverheadReserveBytes = 768ULL << 20;

using metal::BufferStorage;
using metal::CommandGraph;
Expand Down