Skip to main content

Unapologetically SM120-only CuTe DSL kernels for NVFP4 GEMM and MoE.

Project description

b12x is an SM120/SM121 CuTe DSL kernel library for (primarily) NVFP4 LLM inference.

It is intentionally narrow. This is not a generic CUDA kernel collection or a full model-serving stack. It does not intend to target any other GPU architectures, including SM100. It is a focused package for a small number of high-performance kernels plus the runtime glue needed to launch them cleanly from sglang/vllm.

Currently supported kernels:

  • NVFP4 fused MoE GEMM
  • NVFP4 dense GEMM
  • BF16/FP8 paged attention
  • Sparse MLA attention (for DSA/NSA only).
pip install b12x

Ask your friendly neighborhood AI agent for further information on how to use this library.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

b12x-0.12.4.tar.gz (456.7 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

b12x-0.12.4-py3-none-any.whl (515.3 kB view details)

Uploaded Python 3

File details

Details for the file b12x-0.12.4.tar.gz.

File metadata

  • Download URL: b12x-0.12.4.tar.gz
  • Upload date:
  • Size: 456.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.4

File hashes

Hashes for b12x-0.12.4.tar.gz
Algorithm Hash digest
SHA256 89b5ec21d33a7550b401a2fe7c00408b9a8edb1df5359fba305273df02f469ca
MD5 d05c3b1139ae9625d1f7064b21189a74
BLAKE2b-256 505d7a6977b0c84e00601f2dd892fe0879f7a5b4e5973ebec5ee601545495a23

See more details on using hashes here.

File details

Details for the file b12x-0.12.4-py3-none-any.whl.

File metadata

  • Download URL: b12x-0.12.4-py3-none-any.whl
  • Upload date:
  • Size: 515.3 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.4

File hashes

Hashes for b12x-0.12.4-py3-none-any.whl
Algorithm Hash digest
SHA256 f0c57bd8b9ef87d3618414675b2027ee269544ded65784a250d265a4ce30458a
MD5 cd7fa6ad4b01fe865218e136474b7569
BLAKE2b-256 9dc78f1a4d074b18801fe7cdb31869f5e9a6ee8e690fa03e388eb2ba17c5318e

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page