In his dissertation, Hepola develops tools and methodologies that simplify the design, programming and optimisation of custom processors based on the open RISC-V architecture. The proposed approach enables automatic generation of both processor hardware and the compiler from a single description. The work also introduces a methodology for extending RISC-V with static multi-issue capabilities while maintaining high energy efficiency.
The research culminated in the tape-out of a custom digital signal processor implemented in 22 nm technology and designed and programmed using the proposed toolset.
The public defence will take place at Tampere University’s Hervanta campus and can also be followed online. The opponent is Dr César Fuguet from INRIA, France, and the custos is Professor Pekka Jääskeläinen.
Read more about the dissertation and public defence at Tampere University
https://www.tuni.fi/en/tau/news-and-events/kari-hepola-custom-tools-make-risc-v-processors-faster-and-easier-design