deepseek-ai / deepseek-ai/DeepEP

[question] about DeviceRuntime::get_arch

Open
#626 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Cuda
Stars
10.1k
Forks
1.4k
Avg merge
4d 1h
Merged PRs (30d)
2

Description

why arch 10.x is specially converted to `100x` instead of compile with original device arch?

```c++
std::string get_arch(const bool& number_only = false,
const bool& support_arch_family = false) {
const auto [major, minor] = get_arch_pair();
if (major == 10 and minor != 1) {
if (number_only)
return "100";
return support_arch_family ? "100f" : "100a";
}
return std::to_string(major * 10 + minor) + (number_only ? "" : "a");
}
```

Contributor guide

No contributing guide indexed for this repository

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.