PositNN: Tapered Precision Deep Learning Inference for the Edge

@inproceedings{Langroudi2018PositNNTP,
  title={PositNN: Tapered Precision Deep Learning Inference for the Edge},
  author={Hamed Fatemi Langroudi and Zachariah Carmichael and John L. Gustafson and Dhireesha Kudithipudi},
  year={2018}
}
The performance of neural networks, especially the currently popular form of deep 1 neural networks, is often limited by the underlying hardware. Computations in 2 deep neural networks are expensive, have large memory footprint, and are power 3 hungry. Conventional reduced-precision numerical formats, such as fixed-point 4 and floating point, cannot accurately represent deep neural network parameters 5 with a nonlinear distribution and small dynamic range. Recently proposed posit 6 numerical… CONTINUE READING

References

Publications referenced by this paper.
SHOWING 1-10 OF 18 REFERENCES

Densely Connected Convolutional Networks

Gao Huang, Zhuang Liu, Laurens van der Maaten, Kilian Q. Weinberger
  • 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2016
VIEW 3 EXCERPTS
HIGHLY INFLUENTIAL

Optimize Deep Convolutional Neural Network with Ternarized Weights and High Accuracy

  • 2019 IEEE Winter Conference on Applications of Computer Vision (WACV)
  • 2018
VIEW 1 EXCERPT

SBNet: Sparse Blocks Network for Fast Inference

  • 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition
  • 2018
VIEW 1 EXCERPT