Replies to post #136525 on Intel Corp (INTC)

[Unkwn]

Perhaps Core M and Cherry Trail characteristics are to close to each other.

So making one core really good is more rewarding than making two different cores.

First, you seem to forget that Core M isn't anywhere as integrated as a typical mobile SoC. It still carries its separate chipset and most likely needs a lot of peripheral ciruitry as any PC platform does. It's more a shrunken PC platform than a mobile one.

Secondly, the rumor still isn't substantial at all. Just concluding from "Brian Krzanich didn't mention Broxton lately" to "Intel is giving up mobile SoCs" is more than a stretch. It's pure noise without any base.

Handwiring is very important and handwiring in two totally different cores makes not that much sense.

I think they can differentiate with Cache-design, number of cores, with or without hyperthreading and size of the ondie GPU.

I disagree there. Laying out by hand makes sense especially for high speed designs. For mobile, where power and density is most important, synthesis makes more sense (and especially is easier to do). The issue with Core X is that it needs quite a lot of space on die obviously. That's good for the high end PC market but not for mobile. Another core makes sense there.

Having said that, Intel could just take their digital design lead with all the tricks they already use in Core X and integrate it into a mobile SoC. I hope Broxton is just that.