On the Implementation of Fixed-point Exponential Function for Machine Learning and Signal Processing Accelerators

The natural exponential function is widely used in modeling many engineering and scientific systems. It is also an integral part of many neural network activation function such as sigmoid, tanh, ELU, RBF etc. Dedicated hardware accelerator and processors are designed for faster execution of such applications. Such accelerators can immensely benefit from an optimal implementation of exponential function. This can be achieved for most applications with the knowledge that the exponential function for a negative domain (R ) is more widely used than the positive domain (R). This paper presents an optimized implementation of exponential function for variable precision fixed point negative input. The implementation presented here significantly reduces the number of multipliers and adders. This is further optimized using mixed world-length implementation for the series expansion. The reduction in area and power consumption is more than 30% and 50% respectively over previous equivalent method.

[1]  Peter Nilsson,et al.  Hardware implementation of the exponential function using Taylor series , 2014, 2014 NORCHIP.

[2]  Jack E. Volder.  The CORDIC Trigonometric Computing Technique , 1959, IRE Trans. Electron. Comput..

[3]  Stephen Marshall,et al.  Activation Functions: Comparison of trends in Practice and Research for Deep Learning , 2018, ArXiv.

[4]  Mikko H. Lipasti,et al.  SECO: A Scalable Accuracy Approximate Exponential Function Via Cross-Layer Optimization , 2019, 2019 IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED).

[5]  M. C Hanumantharaju,et al.  A Novel Method for Computing Exponential Function Using CORDIC Algorithm , 2012 .

[6]  Doo Seok Jeong,et al.  TS-EFA: Resource-efficient High-precision Approximation of Exponential Functions Based on Template-scaling Method , 2020, 2020 21st International Symposium on Quality Electronic Design (ISQED).

[7]  Ping Tak Peter Tang,et al.  Table-lookup algorithms for elementary functions and their error analysis , 1991, [1991] Proceedings 10th IEEE Symposium on Computer Arithmetic.

[8]  Peter Nilsson,et al.  A VLSI implementation of logarithmic and exponential functions using a novel parabolic synthesis methodology compared to the CORDIC algorithm , 2011, 2011 20th European Conference on Circuit Theory and Design (ECCTD).

[9]  Steve B. Furber,et al.  A fixed point exponential function accelerator for a neuromorphic many-core system , 2017, 2017 IEEE International Symposium on Circuits and Systems (ISCAS).