Search results

Inquiry about ROCm/AMD GPU Support Status

Status: Open.

#1678 In flashinfer-ai/flashinfer;

· Flink-ddd opened on Sep 15, 2025

How to build and install FlashInfer on CUDA 11.8

Status: Open.

#1673 In flashinfer-ai/flashinfer;

· MIKE-GUO233 opened on Sep 11, 2025

Can not reproduce <code>allow_fp16_qk_reduction</code> result in blog

Status: Open.

#1672 In flashinfer-ai/flashinfer;

· cjld opened on Sep 11, 2025

Error on H800

Status: Open.

#1671 In flashinfer-ai/flashinfer;

· shushanxingzhe opened on Sep 11, 2025

DeepSeek FP8 Block Scaling is not yet implemented in CUTLASS fused_moe

priority: high

Status: Open.

#1669 In flashinfer-ai/flashinfer;

· renfeier opened on Sep 11, 2025

BatchPrefillWithRaggedKVCacheWrapper cutlass backend doesn't match FA2 when head_dim=192

Status: Open.

#1663 In flashinfer-ai/flashinfer;

· wenscarl opened on Sep 9, 2025

nvcc fatal : Unsupported gpu architecture 'compute_90a'

Status: Open.

#1658 In flashinfer-ai/flashinfer;

· XingyuLu206 opened on Sep 9, 2025

Feature Request: <code>last_page_len</code> dtype validation

Status: Open.

#1654 In flashinfer-ai/flashinfer;

· roikoren755 opened on Sep 8, 2025

Please don’t give up on SM75.

Status: Open.

#1648 In flashinfer-ai/flashinfer;

· wang824892540 opened on Sep 7, 2025

Error "_vector_sparse_indices_buffer is not large enough" for VariableBlockSparseAttention

Status: Open.

#1647 In flashinfer-ai/flashinfer;

· KevinZeng08 opened on Sep 6, 2025

Using cutlass_fused_moe get NaNs in tensor

Status: Open.

#1646 In flashinfer-ai/flashinfer;

· Oneal65 opened on Sep 6, 2025

[vLLM] Quality Regression with FlashInfer vs Triton Attention

Status: Open.

#1645 In flashinfer-ai/flashinfer;

· jwfromm opened on Sep 5, 2025