In this thesis, I discuss the development of a novel physical design flow introducing standard-cell neurons for ASIC design. Standard-cell neurons are implemented on silicon as a circuit that realizes a threshold function. Each cell contains flash transistors, the threshold voltages of which correspond to the weights of the threshold function. Since the threshold voltages are programmed after fabrication, any sequential logic containing a standard-cell neuron is a logical black box upon delivery to the foundry. Additionally, previous research has shown significant reductions in delay, power, and area with the utilization of these flash transistor (FTL) cells. This paper aims to reinforce this prior research by demonstrating the first automatically synthesized, placed, and routed secure RISC-V core.
Details
- Reducing Delay, Power, and Area of a RISC-V Core Using Standard Cell Neurons
- Grier, Willem (Author)
- Vrudhula, Sarma (Thesis director)
- Singh, Gian (Committee member)
- Barrett, The Honors College (Contributor)
- Computer Science and Engineering Program (Contributor)
- Dean, W.P. Carey School of Business (Contributor)