My lack of follow-up here is indicative of the response I've...

My lack of follow-up here is indicative of the response I've received. I did not hear anything from Peter and then followed up with a similar email to Bob. I did not receive a response to that either. A couple weeks ago I used the website contact form to see if I could get access to the Akida Development Environment. This got a response from Bob asking for more details on my plans. I said that I did not have any immediate plans to incorporate machine learning into any of my projects but that it was a future possibility and that I was interested in learning more about how a developer would go about integrating a neuromorphic chip into a design (which is all true). That was 12 days ago and I didn't hear anything back, so I sent a follow-up email again last Friday.

Regarding the concerns I previously raised, I still have not seen any kind of explanation for the advertised size of the chip and there doesn't appear to be much effort to qualify it; on several occasions the numbers have been given without any qualification. I am confident in the assertion that the claim of 1.2 million neurons and 10 billion synapses is absolutely ludicrous if interpreted as the number of synapses and neurons that are physically implemented on the ASIC according to their published design. So I still find it unsettling that there is no explanation of these numbers, though I also find it completely plausible that there is some sort of interpretation which would make sense (e.g., based on time division multiplexing, unique possible values for registers, etc...). It would just be nice if they would actually offer that explanation somewhere.