Skip to content
View vincentzed's full-sized avatar
🍉
🍉
28 followers · 93 following

Block or report vincentzed

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse

Popular repositories Loading

  1. gemm3 gemm3 Public

    GEMM³: NVFP4 GEMM for NVIDIA B300 / GB300 in CuTe DSL

    Python 3

  2. decon decon Public

    Forked from allenai/decon

    `decon`, but with python API binding.

    Rust 2

  3. smem-swizzle-demo smem-swizzle-demo Public

    Shared-memory swizzling on Blackwell, measured: CuTe C++ & CuTe DSL benches (float4 + ldmatrix), ncu bank-conflict counters, CUPTI timing — 7.8x from one XOR on a B300

    Python 2

  4. cs336-a1-sol cs336-a1-sol Public

    Speed-of-light study on the CS336 A1 45-min OWT leaderboard task: 2.9708 val@512 on B200 (best-ever entry: 3.0354). A measurement of the ceiling, not an entry.

    Python 1

  5. vincentzed vincentzed Public

    GPU inference and model serving.

    1

  6. flashinfer-iket-lab flashinfer-iket-lab Public

    IKET (CuTe-DSL in-kernel event tracing) applied to flashinfer kernels on B300: rmsnorm+fp4quant, NVFP4 blockscaled GEMM, fused GEMM+two-shot allreduce (8 GPU)

    Python 1