Versatile Stochastic Dot-Product Circuit
Tech ID: 30439 / UC Case 2019-427-0
A compact, fast, energy-efficient, and scalable stochastic dot-product circuit.
There have been many demonstrations of efficient hardware for dot product computation, which is the most common operation in stochastic neural networks. In addition, there are numerous reports of the capabilities of stochastic neurons as a feature of Boltzmann machines, which have become the state-of-the-art solution in neural network applications. An efficient and scalable hardware solution combining both functionalities would increase efficiency and performance in the vital spaces of machine learning, information theory, and statistics.
Researchers at the University of California, Santa Barbara have developed a compact, fast, energy-efficient, and scalable stochastic dot-product circuit for neurocomputing and neurooptimization. This circuit is based on two types of memory devices – metal oxide memristors and embedded floating gate memories. Through mixed-signal implementation, efficient stochastic operation is achieved by utilizing the circuit’s noise, which can be intrinsic and/or extrinsic to the array of memory cells. This stochastic dot-product circuit is useful for many applications such as solving optimization problems and developing probabilistic neural networks.
- Increased efficiency and performance
- Compact design
- Fast operation
- Scalable implementation
- Solving optimization problems
- Developing probabilistic neural networks
- VLSI algorithms