2020 BRN Discussion, page-21401

I left out the ultimate point about the implausibility of the original NASA contract with Tensor Partners being unilaterally transformed into one using Intel's Loihi and that is the fact that the Tensor Partners application was approved under SWaP. This program as you know is run by both NASA and DARPA for the purpose of continuously reviewing and replacing existing technology with new technology that reduces Size, reduces Weight and reduces Power and lowers the price of the cost of doing the same thing. The AKD1000 chip in COTS form (commercial off the shelf) at $15.00 and with the power to weight ratio in a 28nm chip blows away anything that Intel currently has to offer. Indeed Intel is already a NASA and DARPA partner and if anyone was going to SWaP neuromorphic technology into NASA's CubeSat program it would be Intel and it would already be happening not left to a small firm of rocket scientists without any actual neuromorphic expertise.

The other thing that I left out is that the NCS2 when you look at the full specs is only rated to operate in a temperature range of 0 Celsius to 40 Celsius and I seem to remember somewhere that it gets a bit colder than this in space and at times considerably hotter so on every level it just is not something that would be capable of interchanging for AKD1000.

My opinion only DYOR.