AKIDA benchmarking?, page-55

A very interesting part of his post was the continual reference back to Nvidia and trying to benchmark Akida against them. My instinct was that this could be an "attack" on behalf of them

Absolutely wrong here. I only mention NVIDIA because they are very well known, and provide benchmarking data on the same standard neural network architectures that are targeted by BrainChip. The Jetson family of devices definitely targets low power applications, but their power consumption is (in principle) way higher than AKIDA. It should be a no-brainer to demonstrate that AKIDA would be the sensible choice for many applications. Brainchip would just need to implement the same problem (using the CNN to SNN convertor), and provide the frame-rate and power consumption information.