Ah, good point, Rick: As BK said, it's all about the DATA. Altera was good, as it's complimentary. Using ARM raised eyebrows, for obviuos reasons, now adding MIPS to the fold makes it even clearer that hardware agnosticism is the way to go. Thanks!
Mobileye's EyeQ architecture is based on MIPS cores. With enough abstraction we can get x86 software to work on MIPS, but that's very inefficient and counterintuitive to what mobile/automotive needs. On the other hand, MIPS is known for its inherent efficiency, though Intel has done well to close the CISC-RISC gap with its micro-ops and many other improvements. So, what becomes of EyeQ 5, that's supposed to debut in 2020? Mobileye's go-to-market strategy has always been its tight coupling of hardware and software for maximum performance and miniumum power.
Will Intel support a MIPS architecture, or will Mobileye have to decouple from MIPS for EyeQ 5? Thoughts? Was this discussed elsewhere?
I think we have the answer here; Intel can't let Nvidia become the ARM of Autos - best move early and move big, as the smartphone experience has demonstrated how expensive it is to be late! Shocked at the price of a ticket to play though, even at this early stage in the market.