A BFP NPU microarchitecture using row/column blocking and per-path protections achieves near-DMR reliability at 3.55% geometric mean performance overhead and under 2% hardware cost.
Flexpoint: An adaptive numerical format for efficient training of deep neural networks.NeurIPS30 (2017)
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
-
From Characterization to Microarchitecture: Designing an Elegant and Reliable BFP-Based NPU
A BFP NPU microarchitecture using row/column blocking and per-path protections achieves near-DMR reliability at 3.55% geometric mean performance overhead and under 2% hardware cost.