Skip to content

issue/632-fix: support iluvatar TG-V200 #633

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

Draft
wants to merge 1 commit into
base: v0.1.0
Choose a base branch
from
Draft

issue/632-fix: support iluvatar TG-V200 #633

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
1 change: 1 addition & 0 deletions src/infiniop/devices/nvidia/nvidia_kernel_common.cuh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,6 +15,7 @@
#define CUDA_BLOCK_SIZE_4096 4096
#define CUDA_BLOCK_SIZE_1024 1024
#define CUDA_BLOCK_SIZE_512 512
#define CUDA_BLOCK_SIZE_2048 2048

#define CHECK_CUDA(API) CHECK_INTERNAL(API, cudaSuccess)

Expand Down
4 changes: 4 additions & 0 deletions src/infiniop/ops/causal_softmax/nvidia/causal_softmax_nvidia.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -88,6 +88,10 @@ infiniStatus_t Descriptor::calculate(void *workspace, size_t workspace_size,
CHECK_STATUS(launchKernel<CUDA_BLOCK_SIZE_4096>(
y, x, _info.dtype, _info.batch_size, _info.seq_len, _info.total_seq_len,
_info.y_stride_b, _info.y_stride_i, _info.x_stride_b, _info.x_stride_i, stream));
} else if (_opaque->internal->maxThreadsPerBlock() == CUDA_BLOCK_SIZE_2048) {
CHECK_STATUS(launchKernel<CUDA_BLOCK_SIZE_2048>(
y, x, _info.dtype, _info.batch_size, _info.seq_len, _info.total_seq_len,
_info.y_stride_b, _info.y_stride_i, _info.x_stride_b, _info.x_stride_i, stream));
} else {
return INFINI_STATUS_DEVICE_ARCHITECTURE_NOT_SUPPORTED;
}
Expand Down
2 changes: 2 additions & 0 deletions src/infiniop/ops/rms_norm/nvidia/rms_norm_nvidia.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -123,6 +123,8 @@ infiniStatus_t Descriptor::calculate(
CHECK_STATUS(launchKernel<CUDA_BLOCK_SIZE_512>(batch_size, nhead, dim, y, _info.atype, stride_y_batch, stride_y_nhead, x, stride_x_batch, stride_x_nhead, w, _info.wtype, _info.epsilon, cuda_stream));
} else if (_opaque->internal->maxThreadsPerBlock() == CUDA_BLOCK_SIZE_4096) {
CHECK_STATUS(launchKernel<CUDA_BLOCK_SIZE_4096>(batch_size, nhead, dim, y, _info.atype, stride_y_batch, stride_y_nhead, x, stride_x_batch, stride_x_nhead, w, _info.wtype, _info.epsilon, cuda_stream));
} else if (_opaque->internal->maxThreadsPerBlock() == CUDA_BLOCK_SIZE_2048) {
CHECK_STATUS(launchKernel<CUDA_BLOCK_SIZE_2048>(batch_size, nhead, dim, y, _info.atype, stride_y_batch, stride_y_nhead, x, stride_x_batch, stride_x_nhead, w, _info.wtype, _info.epsilon, cuda_stream));
} else {
return INFINI_STATUS_DEVICE_ARCHITECTURE_NOT_SUPPORTED;
}
Expand Down
2 changes: 1 addition & 1 deletion xmake/iluvatar.lua
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -43,7 +43,7 @@ target("infiniop-iluvatar")

set_warnings("all", "error")
add_cuflags("-Wno-error=unused-private-field")
add_cuflags("-fPIC", "-x", "ivcore", "-std=c++17", {force = true})
add_cuflags("-fPIC", "-x", "ivcore", "-std=c++17", "--cuda-gpu-arch=ivcore20", {force = true})
Copy link
Collaborator

@PanZezhong1725 PanZezhong1725 Nov 24, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

这么改旧版还支持吗？

add_culdflags("-fPIC")
add_cxflags("-fPIC")

Expand Down