Seven Nanometer, Ten Nanometer, Fourteen Nanometer discussed on PC Perspective Podcast

Automatic TRANSCRIPT

And would you say that it is overperforming underperforming i i have no idea but it's a two core not gpu part so chances are as it's not it's not it's not good there are two trains of thought here one ten nanometer is it can be fine the things we just gotta work out and tweak it and everything's fine the other are it's fundamentally unsound and you will never get it to work at the levels you needed to be i i am of two minds of this i think in its current configuration it probably will never meet the benchmarks they needed to but there are many things that they can do because you know process analogy is is like science and alchemy combined it's really absolutely bizarre until until they see how something works and can quantify it it's like they don't really know if if it will work or not i mean there's theory and obviously theory to reality for intel broke down somewhere that they need to take the reality and make adjustments and see what comes out and i you know everybody can be critic because it's easy to say you done screwed up but yeah i don't know where that they really are going to go i mean they they now have two options one fixed ten nanometer or scrap entirely and just focus on seven nanometer and try to learn the lessons that they got from ten nanometer i mean they threw a bunch of technologies at ten animator i mean cobalt i can't remember all the other different you know gate features that that they had sean but it just it was a lot of stuff and it was all in a very very tight package and if you look at the physical dimensions ten then a meter is still tighter overall than the competition seven nanometer parts like tsmc but if they can't make it it doesn't matter it doesn't matter because you know if you crap and he products you make are extremely simple the people who can leverage twelve nanometer which is you know samsa samsung scoble foundries global foundries you know enhanced fourteen nanometer stuff if you can put out parts at will and still be able to compete with intel at their fourteen plus plus then tells losing money there they do not have command of the the market that they previously did because essentially the playing level the playing field is level right now and it's going to get worse for intel in twenty nineteen so yeah they've made a boatload of money in twenty seventeen and twenty team twenty eighteen is going to close out very very strong for them as they still have very good parts out there they are they're keeping their margins they still have all the customers that they have but twenty nineteen is is going to be an inflection point because amd will have seven nanometer parts intel will not have ten nanometer stuff they will still have to rely on fourteen animator plus plus plus plus plus whatever they decide do amdi is going to have a density advantage over intel and that is going to be huge for data center i don't think that amdi is going to take over data center but attentively they could take up words of ten percent of intel's really high margin high end big money data center with their epic processors especially seven made him so let's not try to guess who will replace him that's too as we just be guessing do you think what changes when you have a successor right and i guess part of this question is impossible to answer until you know who the successor is do we see any shift in direction for what they're doing right if you look they have where the rat on the desktop cpu cyber on service ep side or the rat on mobile where they're at on a i do they pga's do they change anything this is this is actually the second big loss for them or removal.

Coming up next