2022 BRN Discussion, page-9766

Exactly...Qualcomm have been researching neuromorphic hardware for quite some time but what edge AI platform have they produced?
They have a fantastic silicon architecture ready to fit Brainchip's IP at some stage...unless ARM does it first...or AMD or Nvidia...the possibilities!
Here is an interesting look at the mindset of Venkat Rangan who was the Director of Engineering at Qualcomm.
https://web.stanford.edu/class/ee380/Abstracts/100331-slides
This was his presentation for the Qualcomm's R&D program possibly 10 years ago...his vision for a neuromorphic chip is exactly what Brainchip has produced! Brainchip's IP have certainly answered most of his proposed challenges!
His last statement "So you have the biggest ever network of neurons/synapses, now what?" is the most telling...at the end of the day, everyone wants the biggest neurons/synapses platform but I would also add that these neurons/synapses need to internally work well with one another (i.e, using JAST!) and externally compatible with any silicon architecture (eg, ARM, Qualcomm) and any algorithm platform (eg, Edge Impulse), plus retrospectively compatible with previous neural network to allow smoother transition to this "next generation" platform.
So to me Brainchip has ticked all these boxes and after a year, no one else has appear to come close. There have been variations of earlier platforms but you would not call these "next generation"!