Design Techniques for Multi-Core Neural Network Accelerators on Radiation-Hardened FPGAs

Radiation-Hardened-By-Design (RHBD) FPGAs have gained a lot of attention thanks to their excellent compromise between costs and performance. Being of very limited use due to a lack of performance a few years ago, these devices are now capable of implementing a wide range of applications requiring high computational capabilities. This work describes an implementation of a Very Long Instruction Word (VLIW) soft-core convolutional accelerator in the NanoXplore RHBD NG-Medium chip. Feasibility and timing performances have been analyzed in order to discover whether and how multi-core solutions can affect parallel acceleration. Placement also showed to heavily affect the delays, up to 70% more, based on the proximity to the output buffers.