A lightweight shared-memory technique for NF4 dequantization kernels yields 2.0-2.2x kernel speedup and 1.54x end-to-end gains on models up to 70B parameters while using only 64 bytes of shared memory per block.
The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Fast NF4 Dequantization Kernels for Large Language Model Inference
A lightweight shared-memory technique for NF4 dequantization kernels yields 2.0-2.2x kernel speedup and 1.54x end-to-end gains on models up to 70B parameters while using only 64 bytes of shared memory per block.