Search results

[issue] st.bulk instruction requires sm_100, don't support pre-sm100 arch

Status: Open.

#675 In deepseek-ai/DeepEP;

· huakainomore opened on Jul 3, 2026

[BUG] DeepEPv2 internode num_worst_tokens tests

Status: Open.

#670 In deepseek-ai/DeepEP;

· GeofferyGeng opened on Jun 23, 2026

[Performance] Question about performance of Deepep v2, the role of the prefer-overlap-with-compute parameter

Status: Open.

#669 In deepseek-ai/DeepEP;

· shenhongqian opened on Jun 22, 2026

Question: Why did DeepEP v2 drop the double-buffer design in favor of explicit pre/post-sync barriers?

Status: Open.

#668 In deepseek-ai/DeepEP;

· mkbk-with-circle opened on Jun 19, 2026

What is the function of check_fast_rdma_atomic_support

Status: Open.

#667 In deepseek-ai/DeepEP;

· shenhongqian opened on Jun 16, 2026

Would deepep work in a hybrid GPU networking topology inside a single node(intronode)?

Status: Open.

#666 In deepseek-ai/DeepEP;

· shrutimohanty opened on Jun 15, 2026

[Question] What's policy about apply ElasticBuffer for prefill?

Status: Open.

#662 In deepseek-ai/DeepEP;

· xiaoqi35 opened on Jun 11, 2026

Can HybridEP NIXL path support EFA/libfabric, or is it UCX/GDA-specific?

Status: Open.

#659 In deepseek-ai/DeepEP;

· irenedea opened on Jun 3, 2026

Does NVSHMEM/DeepEP support SR-IOV Virtual Function devices?

Status: Open.

#658 In deepseek-ai/DeepEP;

· hmthhh opened on Jun 3, 2026

[DeepEPv2] test_agrs.py failed due to CUDA_ERROR_NOT_SUPPORTED, operation not supported

Status: Open.

#656 In deepseek-ai/DeepEP;

· elevenxiang opened on Jun 1, 2026

Zero-copy intranode for EPv2

Status: Open.

#655 In deepseek-ai/DeepEP;

· usernamehaha2022 opened on Jun 1, 2026

[Performance] Question About Dispatch Prologue Barrier and Excessive P99 ITL With V2

Status: Open.

#652 In deepseek-ai/DeepEP;

· elvircrn opened on May 30, 2026