CRAM-ER combines spintronic computational RAM with CMOS adder trees and software fine-tuning to deliver near-lossless DNN accuracy at up to 100x lower latency than CPU/GPU baselines.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
verdicts
UNVERDICTED 2representative citing papers
Multiple samples from probabilistic networks improve DNN accuracy over single deterministic samples, with an energy tradeoff formula for choosing between more samples and more bits.
citing papers explorer
-
CRAM-ER: Error-Resilient Spintronic Computational Random Access Memory for Scalable In-Memory Computation
CRAM-ER combines spintronic computational RAM with CMOS adder trees and software fine-tuning to deliver near-lossless DNN accuracy at up to 100x lower latency than CPU/GPU baselines.
-
Improving deep neural network performance through sampling
Multiple samples from probabilistic networks improve DNN accuracy over single deterministic samples, with an energy tradeoff formula for choosing between more samples and more bits.