All issues

Search results

[RFC] Fused W8A8 MoE megakernel on H20

Status: Open.

#94 In Tencent/hpc-ops;

· Mikezhang001 opened on Aug 28, 2026

Integration for GLM-4.6-FP8 and performance tune on H20

Status: Open.

#80 In Tencent/hpc-ops;

· lichang98 opened on Aug 4, 2026

[Question] Support head_dim=256 for attention ops (Qwen3.5/3.6-style GDN-hybrid models) ？

Status: Open.

#70 In Tencent/hpc-ops;

· yaofeino1 opened on Jul 21, 2026

[Question] How to Understand the thr_layout and val_layout in make_tiled_copy_V_interleave_trans?

Status: Open.

#44 In Tencent/hpc-ops;

· Maximilianxu opened on May 18, 2026

[Feature request] decode attention support mtp-3

Status: Open.

#43 In Tencent/hpc-ops;

· ZelinMa557 opened on May 17, 2026

Potential Shared Memory Misalignment Issues in GEMM Kernels

Status: Open.

#42 In Tencent/hpc-ops;

· trobr opened on Apr 13, 2026

Any plan to support DeepSeek V3.2 ?

Status: Open.

#33 In Tencent/hpc-ops;

· lichang98 opened on Mar 16, 2026

Ada fp8 kernel support

Status: Open.

#31 In Tencent/hpc-ops;

· starwang1024 opened on Feb 26, 2026

[Performance] GroupedGEMM seems not to be faster than DeepGEMM

Status: Open.

#28 In Tencent/hpc-ops;

· NiuMa-1234 opened on Feb 11, 2026

FP4 / NVFP4 / MXFP4 support

Status: Open.

#25 In Tencent/hpc-ops;

· arcticgus opened on Feb 5, 2026

test_fused_moe_blockwise.py 改了一下shape，单测过不了

Status: Open.

#19 In Tencent/hpc-ops;

· K11OntheBoat opened on Feb 2, 2026

Sglang integrations

Status: Open.

#16 In Tencent/hpc-ops;

· Kamizm opened on Jan 30, 2026