Sasikanth Avancha — accepted AI-conference papers

2020

dMazeRunner: Optimizing Convolutions on Dataflow Accelerators

Shail Dave, Aviral Shrivastava, Youngbin Kim, Sasikanth Avancha, Kyoungwoo Lee

ICASSP 2020accepted

Convolution neural networks (CNNs) can be efficiently executed on dataflow accelerators. However, the vast space of executing convolutions on computational and memory resources of accelerators makes difficult for programmers to automatically and efficiently accelerate the convolutions and for archit…

Cited by 0 Source ↗Scholar ↗

2018

Mixed Precision Training of Convolutional Neural Networks using Integer Operations

Dipankar Das, Naveen Mellempudi, Dheevatsa Mudigere, Dhiraj Kalamkar, Sasikanth Avancha, Kunal Banerjee +11

ICLR 2018poster

The state-of-the-art (SOTA) for mixed precision training is dominated by variants of low precision floating point operations, and in particular, FP16 accumulating into FP32 Micikevicius et al. (2017). On the other hand, while a lot of research has also happened in the domain of low and mixed-precisi…

Cited by 226 Source ↗PDF ↗Scholar ↗