Hardware Accelerator for Solving Differential Equations
This project involves the development of a hardware accelerator module to solve the Lotka-Volterra family of differential equations, which are used in the field of computational biology. The accelerator was developed using the Chisel language and the Chipyard environment, which includes RISC-V CPUs and other hardware accelerators written in Chisel.
The purpose of a hardware accelerator is to increase performance and decrease energy consumption for a specific type of computation, in this case solving a specific family of differential equations. The equations are implemented directly in hardware, creating an accelerator module that interfaces with a general-purpose RISC-V CPU.
Currently the accelerator works properly for the basic set of Lotka-Volterra equations, though the performance is not great. This is still very much a work in progress, and its development will continue.
The software I wrote for this can be found at the GitLab repository.