Communication about changes to the company's business plans, page-15

Understand your concern, I see it this way.

From my thinking path to market and $$$ is accelerated by the new model of BRN SNN on FPGA instead of onchip ASIC. M4 can be achieved much more quickly with this model, I hope.

Microsoft /Azure & Amazon / AWS EC2

are looking at FPGA for their AI and Deep learning and Neural Net workloads as opposed to Google who are looking at Tensor Processors for same workloads. FPGA are more easily changed and programmable wheres dedicated ASIC require to be replaced.

I see advantage of BRN SNN on FPGA first as this can be more easily trialed for particular workloads and applications in cloud as opposed to a dedicated ASIC. If BRN SNN on FPGA can be proven to offer significant speed advantage this can be taken up with servers utilising Microsoft Project Olympus and specific AI workloads can be run in private cloud.

Long term ASIC can be pursued but FPGA in my opionion is the smarter short term move. Once proven IP can also be second wave to market as it were to chip makers or the like via the ARM model?? If that is pursued.

refs:

https://www.top500.org/news/microsofts-plans-for-fpgas-in-azure-should-worry-traditional-chipmakers/

https://techcrunch.com/2016/11/30/aws-announces-fpga-instances-for-its-ec2-cloud-computing-service/

Microsoft Project Olympus FGPA

https://azure.microsoft.com/en-us/b...pport-for-new-intel-xeon-scalable-processors/

& Nice reading below on the landscape of AI in relation to FPGA and ASIC for investors

https://seekingalpha.com/article/4085429-nvidia-choosing-sit-one

IMOH/ DYOR / GLTAH yadda yadda