Skip to content

vulkan: use CPU writes in ggml_backend_vk_cpy_tensor_async if the context is idle #28618

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

Merged
ggerganov merged 1 commit into from
Sep 10, 2026
Merged

vulkan: use CPU writes in ggml_backend_vk_cpy_tensor_async if the context is idle #28618

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
17 changes: 16 additions & 1 deletion ggml/src/ggml-vulkan/ggml-vulkan.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16998,6 +16998,22 @@ static bool ggml_backend_vk_cpy_tensor_async(ggml_backend_t backend_src, ggml_ba
return false;
}

// If the backend is idle, use a CPU copy to avoid GPU synchronization overhead.
static constexpr size_t max_cpu_copy_size = 128 * 1024;
const bool src_backend_synchronous = backend_src->iface.synchronize == nullptr;
const bool transfer_idle = !ctx->device->async_use_transfer_queue ||
ctx->transfer_semaphore_last_submitted == ctx->transfer_semaphore.value;
const bool backend_idle = ctx->compute_ctx.expired() && ctx->transfer_ctx.expired() &&
!ctx->submit_pending && !ctx->almost_ready_fence_pending && transfer_idle;
const bool dst_host_coherent =
(dst_buf->memory_property_flags & (vk::MemoryPropertyFlagBits::eHostVisible | vk::MemoryPropertyFlagBits::eHostCoherent)) ==
(vk::MemoryPropertyFlagBits::eHostVisible | vk::MemoryPropertyFlagBits::eHostCoherent);

if ((backend_src == backend_dst || src_backend_synchronous) && backend_idle && dst_host_coherent && ggml_nbytes(src) <= max_cpu_copy_size) {
ggml_vk_buffer_write(dst_buf, vk_tensor_offset(dst) + dst->view_offs, src->data, ggml_nbytes(src));
return true;
}

vk_context cpy_ctx;
if (ctx->device->async_use_transfer_queue) {
cpy_ctx = ggml_vk_get_transfer_ctx(ctx);
Expand All @@ -17010,7 +17026,6 @@ static bool ggml_backend_vk_cpy_tensor_async(ggml_backend_t backend_src, ggml_ba
src->data, ggml_nbytes(src));
}

GGML_UNUSED(backend_src);
return false;
}

Expand Down
Loading