Popular repositories Loading

1. vLLM-Moet vLLM-Moet Public

   A vLLM patch + hand‑written SM120 SASS kernels: 2‑bit MoE experts + an FP4 "delta" cache that recovers precision — matching the official (NV)FP4 checkpoint's quality on consumer Blackwell cards

   Sass 490 44

2. qwentin qwentin Public

   27B at 256k context on one RTX 5090 — FP6 + 4-bit KV + MTP spec-decode, hand-written SM120 tensor-core kernels, OpenAI API.

   Cuda 11

3. LLobotomy LLobotomy Public

   Surgical uncensoring of large language models via Optimal Transport activation hooks

   Python 6

4. cubit cubit Public

   Open-source SM120 (Blackwell) SASS assembler — encode, schedule & assemble GPU machine code, including all tensor-core instructions.

   Rust 6 1

5. rtx-p2p rtx-p2p Public

   PCIe P2P on consumer RTX 5090 via one GSP firmware registry key — no driver patches.

   Shell 4 4

6. blackwell-isa blackwell-isa Public

   Reverse-engineered instruction set architecture database for NVIDIA SM120 (Blackwell) GPUs.

   HTML 3