LUT-DLA converts neural network layers into vector-quantized lookup tables, claiming sub-1-bit-equivalent inference with 1.4-7.0x power and 1.5-146.1x area efficiency gains over conventional accelerators.
Pqa: Exploring the potential of product quantization in dnn hardware acceleration,
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.AR 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
LUT-DLA: Lookup Table as Efficient Extreme Low-Bit Deep Learning Accelerator
LUT-DLA converts neural network layers into vector-quantized lookup tables, claiming sub-1-bit-equivalent inference with 1.4-7.0x power and 1.5-146.1x area efficiency gains over conventional accelerators.