SGLang Quantization Documentation
repo
Fields
id
sources/qhw-src-sglang-quant-docsPublisher
SGLang
Kind
repo
Accessed
2026-10-04
notes
SGLang quantization documentation. Contains platform compatibility table for quantization methods across NVIDIA, AMD, and Ascend hardware. Includes GEMM backend selection logic for FP4/FP8 quantization with auto-selection order. Documents kernel selection based on hardware compute capability (SM80, SM90, SM100, SM120).
Notes
SGLang quantization documentation. Contains platform compatibility table for quantization methods across NVIDIA, AMD, and Ascend hardware. Includes GEMM backend selection logic for FP4/FP8 quantization with auto-selection order. Documents kernel selection based on hardware compute capability (SM80, SM90, SM100, SM120).
Cross-references
This record references no other records.