> would be interesting to see them make a mobile part designed for phones.
I worked on a well-funded effort to build a top-of-the-line mobile AP (the Broxton half of these [two cancelled programs]). It was a shit show from start to finish. There is so much mismatch between Intel's strengths and what the industry demands that I don't see Intel being successful in this domain if they make another attempt. Things would be worse now than in that attempt, as Intel has suffered significant brain-drain since then.
Intel has been behind on process before and come out fine. They've been ahead on process and technology, but at those times should have been able to capitalize on it to get into mobile and GPUs early on I feel. They couldn't/didn't do it.
I was eagerly awaiting Broxton & successors, and now FPGAs and Optane just aren't going to be as big as GPUs and mobile. They have no choice but to keep investing in GPUs and hope for the best in CPUs.
x86 works in mobile form factors, as evidenced by the Silvermont-based SoCs that made it into a few phones, including a reasonably successful Zenphone.
The problem I saw was that the design philosophy at Intel was not conducive to fast turnaround, and the mobile market does not wait.
I worked on a well-funded effort to build a top-of-the-line mobile AP (the Broxton half of these [two cancelled programs]). It was a shit show from start to finish. There is so much mismatch between Intel's strengths and what the industry demands that I don't see Intel being successful in this domain if they make another attempt. Things would be worse now than in that attempt, as Intel has suffered significant brain-drain since then.
[two cancelled programs]:https://www.anandtech.com/show/10288/intel-broxton-sofia-sma...