CARMEN delivers a CORDIC-based multi-precision vector engine achieving up to 33% fewer computation cycles and 21% power savings per MAC in 28 nm ASIC while supporting 8/16-bit flexible precision for deep learning inference.
An Empirical Approach to Enhance Performance for Scalable CORDIC-Based DNNs,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
CARMEN: CORDIC-Accelerated Resource-Efficient Multi-Precision Inference Engine for Deep Learning
CARMEN delivers a CORDIC-based multi-precision vector engine achieving up to 33% fewer computation cycles and 21% power savings per MAC in 28 nm ASIC while supporting 8/16-bit flexible precision for deep learning inference.