paper-with-me

홈 › Papers

Cnvlutin2: Ineffectual-Activation-and-Weight-Free Deep Neural Network Computing

2017-04-29 · Patrick Judd, Alberto Delmas, Sayeh Sharify, Andreas Moshovos

We discuss several modifications and extensions over the previous proposed Cnvlutin (CNV) accelerator for convolutional and fully-connected layers of Deep Learning Network. We first describe different encodings of the activations that are deemed ineffectual. The encodings have different memory overhead and energy characteristics. We propose using a level of indirection when accessing activations from memory to reduce their memory footprint by storing only the effectual activations. We also present a modified organization that detects the activations that are deemed as ineffectual while fetching them from memory. This is different than the original design that instead detected them at the output of the preceding layer. Finally, we present an extended CNV that can also skip ineffectual weights.

📄 PDF Abstract BibTeX arXiv:1705.00125

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Bit-Tactical: Exploiting Ineffectual Computations in Convolutional Neural Networks: Which, Why, and How

2018-03-09 · Alberto Delmas, Patrick Judd, Dylan Malone Stuart, Zissis Poulos 외

We show that, during inference with Convolutional Neural Networks (CNNs), more than 2x to $8x ineffectual work can be exposed if instead of targeting those weights and activations that are zero, we target different combi…

Tetris: Re-architecting Convolutional Neural Network Computation for Machine Learning Accelerators

2018-11-14 · Hang Lu, Xin Wei, Ning Lin, Guihai Yan 외

Inference efficiency is the predominant consideration in designing deep learning accelerators. Previous work mainly focuses on skipping zero values to deal with remarkable ineffectual computation, while zero bits in non-…

BIG-bench Machine Learning

Laconic Deep Learning Computing

2018-05-10 · Sayeh Sharify, Mostafa Mahmoud, Alberto Delmas Lascorz, Milos Nikolic 외

We motivate a method for transparently identifying ineffectual computations in unmodified Deep Learning models and without affecting accuracy. Specifically, we show that if we decompose multiplications down to the bit le…

2kDeep Learningimage-classificationImage Classification

Bit-pragmatic Deep Neural Network Computing

2016-10-20 · J. Albericio, P. Judd, A. Delmás, S. Sharify 외

We quantify a source of ineffectual computations when processing the multiplications of the convolutional layers in Deep Neural Networks (DNNs) and propose Pragmatic (PRA), an architecture that exploits it improving perf…

Efficiently Deploying LLMs with Controlled Risk

2024-10-03 · Michael J. Zellinger, Matt Thomson

Deploying large language models in production requires simultaneous attention to efficiency and risk control. Prior work has shown the possibility to cut costs while maintaining similar accuracy, but has neglected to foc…

MMLUTruthfulQA