Repository Issue Activity (beta)

#356 Does Mega MoE support RTX 50-series GPUs, and can it work in a PCIe-only setup without NVLink?

#322 `get_paged_mqa_logits_metadata()` does not work with large batch size due to excessive static shared memory impl

#232 [Feature Request] Pip install support

#339 [BUG] Mega MoE kernel中的小错误

#6 Support for Ada Lovelace & Blackwell

#290 [QUESTION]Why does fp8_mqa_logits use four MATH WGs for computation?

#338 [RFC] Support FP4 (E2M1) activation dispatch in MegaMoE

#336 Question about Nsight Compute results for sm100_fp8_fp4_mega_moe_impl on B300/Blackwell

#335 Will DeepGEMM support GEMM + 1d1d quantization all in one fused kernel?

#334 [Feature Request] Roadmap for Ascend NPU support in MegaMoE?

#329 最新的mega_moe算子，支持nvfp4格式吗，或者有计划支持nvfp4格式吗？

#333 [Feature Request] Make binary artifacts portable on torch and Python versions

#330 [SM100/B300][BF16] bf16_gemm_nt failure with prebuilt wheel, fixed by rebuilding against current PyTorch

#320 [Question]About the new maga kernel and swapab feature

#306 About the version of DeepEP for mega_moe_kernel

#305 need support sm120

#321 question about megamoe

#317 DeepSeek-V4 on SM120: tf32_hc_prenorm_gemm and paged_mqa_logits kernels missing

#228 __version__ doesn't match Git tags for post-releases, breaking build tools

#236 Feature Request: Support sm_120 ( 5090 and blackwell 6000 pro )

#208 Does DeepSeek V3.2 indexer logit kernels have an API for implementing bfloat16

#199 The comparison between cublaslt and deepgemm in test_fp8.py is unaligned

#315 [Question] Do you have any plan to impl mega_mla_kernel or mega_mla_with_indexer_kernel?

#313 [Question] Is there a simpler implementation for the UTCCP transposer?

#309 fix: operator precedence bug in pack_ue8m0_to_int assertion (mantissa check always passes)