Design and Analysis of a Hardware CNN Accelerator

@inproceedings{Kiningham2017DesignAA,
  title={Design and Analysis of a Hardware CNN Accelerator},
  author={Kevin Kiningham},
  year={2017}
}
In recent years, Convolutional Neural Networks (CNNs) have revolutionized computer vision tasks. However, inference in current CNN designs is extremely computationally intensive. This has lead to an explosion of new accelerator architectures designed to reduce power consumption and latency [20]. In this paper, we design and implement a systolic array based architecture we call ConvAU to efficiently accelerate dense matrix multiplication operations in CNNs. We also train an 8-bit quantized… CONTINUE READING

References

Publications referenced by this paper.
SHOWING 1-10 OF 26 REFERENCES

and K

F. N. Iandola, S. Han, M. W. Moskewicz, K. Ashraf, W. J. Dally
  • Keutzer. Squeezenet: Alexnet-level accuracy with 50x fewer parameters and¡ 0.5 mb model size. arXiv preprint arXiv:1602.07360
  • 2016
VIEW 5 EXCERPTS
HIGHLY INFLUENTIAL

et al

N. P. Jouppi, C. Young, +7 authors A. Borchers
  • In-datacenter performance analysis of a tensor processing unit. arXiv preprint arXiv:1704.04760
  • 2017
VIEW 6 EXCERPTS
HIGHLY INFLUENTIAL

ImageNet Large Scale Visual Recognition Challenge

  • International Journal of Computer Vision
  • 2014
VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

24.2 A 2.5GHz 7.7TOPS/W switched-capacitor matrix multiplier with co-designed local memory in 40nm

  • 2016 IEEE International Solid-State Circuits Conference (ISSCC)
  • 2016
VIEW 1 EXCERPT

ISAAC: A Convolutional Neural Network Accelerator with In-Situ Analog Arithmetic in Crossbars

  • 2016 ACM/IEEE 43rd Annual International Symposium on Computer Architecture (ISCA)
  • 2016
VIEW 1 EXCERPT