https://edgeimpulse.com/blog/brainchip-akida-and-edge-impulseTher...

https://edgeimpulse.com/blog/brainchip-akida-and-edge-impulse


There is no doubt about Brainchip's neuromorphic IP...just ask VVDN, EDGX, Teksun who recently came out with edge AI solutions using Akida.

https://venturebeat.com/ai/neuromorphic-computing-the-long-path-from-roots-to-real-life/

This article gives a great perspective regarding the path of neuromorphic computing, its benchmark with current CPU/GPU computers and how it will complement CPU, GPU, and FPGA technologies for certain tasks (such as learning, searching and sensing), with extremely low power and high efficiency.

https://www.researchgate.net/publication/372654514_Implementing_and_Benchmarking_the_Locally_Competitive_Algorithm_on_the_Loihi_2_Neuromorphic_Processor

This is the latest benchmarking of Loihi 2...impressive. We all know that Akida 2 will benchmark just as well if not better than Loihi 2...will Intel ever perform this test?...I doubt it!...will Brainchip ever perform this test?...maybe.

https://www.anandtech.com/show/16960/intel-loihi-2-intel-4nm-4
https://www.reuters.com/technology/intel-says-newest-laptop-chips-software-will-handle-generative-ai-2023-09-19/

These 2 articles outline Intel's AI chip plan...first introduce the "AI PC" with its latest CPU with an added NPU, then likely introduce Loihi 2 to work with this "AI PC". This is when I think neuromorphic computing will have its day and the ARM/Akida will be an alternative to Loihi 2...industries always develop much faster with competition.
One interesting fact that I found out which explains ARM's affinity with Akida: "The University of Manchester worked to tackle neural algorithms on low-power hardware with its Spiking Neural Network Architecture (SpiNNaker) supercomputer, built from 57,600 processing nodes, each of eighteen 200 MHz ARM 9 processors. The SpiNNaker project spotlights a particularly critical problem in this space: Despite using ARM processors, the solution still spans 10 rack-mounted blade enclosures and requires roughly 100 kW to operate. Learning systems in edge-based applications don’t have the liberty of such power budgets." So the ARM/Akida chip will be the SpiNNaker project's dream child!

https://asia.nikkei.com/Business/Tech/Semiconductors/Apple-Samsung-to-invest-in-Arm-as-it-eyes-September-IPO

"Leading global chipmakers, including Apple, Samsung Electronics, Nvidia and Intel, will invest in Arm as soon as the company is listed on the market." This is very interesting.
I remember reading somewhere that Brainchip's next goal is to have increased "in-memory" processing with its SNN which is quite low at the moment. What is this all about? Currently, Akida can only learn, infer and filter real-time sensory information to more central CPU/GPU computers with all the memory to produce an output such as perform a task or graphics. When Akida has sufficient "in-memory" it can then produce "outputs" without always the need for CPU/GPU computers.
https://hothardware.com/news/samsung-brain-like-mram-tech-claims-to-excel-at-ai-workloads
Samsung has been working hard on its "in-memory" solution ie, MRAM.
Collaboration between the ARM/Akida chip with Samsung's MRAM for instance may make a truly powerful neuromorphic chip that could do away with CPU/GPU computers and I think this will happen in the next year or so. This will mean you can literally have a mobile or watch that is as powerful as a modern day computer that can also learn and process sensory information on the fly and produce the tasks and graphics (maybe holographics) as required.

This is my big picture of the current AI landscape outside the cloud that Brainchip will be in the mix. As for the cloud, I think quantum computing will takeover the thousands of CPU/GPU needed to run the cloud...although neuromorphic computing may bridge this gap for a period of time.
I truly have no idea what the SP will do and Brainchip is quite vulnerable at the moment for all sorts of manipulation and geopolitical tensions doesn't help. To be honest, Peter and Anil should have kept it as a private company but then again the little investors like ourselves will never get the opportunity to own a part of history...possibly that was their plan so that no one man can really control AI !