Straw man much? I didn't imply that Intel is paying off Toms, nor even that Toms is biased. I said "calling in favors", and that's exactly what such an uncritical re-posting of extravagant PR puffery looks like. It's very likely to mislead less-informed readers - which of course is Intel's intent. Toms needs to draw the line somewhere, and I think they missed that here.
And - it's TSMC's new process, not Samsung... you'll want to have that fact correct, going forward.
Maybe not but the way you stated it, and others are, allude to it.
But hey its fine to basically post the information from AMD but not Intel? They are literally just reporting the information given. They do it for every major PR release from major companies. It doesn't mean the guy writing it is an Intel fanboy. Its just basic reporting.
They didn't miss any line anywhere. The article that missed the line was the "Just Buy It" article about nVidias RTX before the card launched. That missed the line. Reporting the information from Computex for both sides is not missing the line. Now if they said "Oh gee gooly, AMD is in for a world of hurt!!! 18% guaranteed gains!!!" that would be different but relaying Intels claims is not missing any line.
And correct. TSMC. I mixed them up but thanks for the correction.
This is a mobile chip, not a desktop chip. Power numbers matter more than peak boost rates.
You mean Samsungs new process?
It's
TSMC , So now we attack the the manufaturer? Is intels 18% increase on patched or unpatched coffelake? This is a 4 year late mobile chip and will be a 5 year late desktop chip, think the desktop chip will need a new socket? or perhaps same socket just a new chipset to use lower power chips
Perhaps they should speak to Samsung or TSMC maybe get a little help, they need it.
Not attacking any manufacture. AMD has not designed a process for quite a while but AMD is in no way responsible for any advancements the 7nm process brings to their CPU, that would be as I was corrected TSMC.
I have no idea how they did the 18%. I would assume unpatched as I would assume the fixed are built into the hardware. However if you look at the first line I said its best to always wait till third party sites get to test the products and that the 18% might only be certain scenarios.It may be late but what does that matter? if it truly is capable of an 18% IPC gain then late or not it doesn't matter.
My assumption would be new socket. Until we see what Intel has planned its all speculation. They plan to have 10nm server next year and I have not seen anything for desktop yet on 10nm but 7nm is supposed o be 2021. We shall see.
And considering that Samsung and TSMC normally work with entire collations on process tech vs Intel working alone I would say Intel did a pretty damn good job at keeping ahead of most of them. I don't think Intel would ask for the help, it might benefit them but they also push things further than Samsung or TSMC. Intels 10nm was going to be quite a bit more dense than TSMC and Samsungs 7nm (might have changed, not sure until we see it) when it was first proposed for a 2015 launch. Intels 7nm will also be more dense than Samsungs 5nm as of the last information I heard about it.
Facebook
Twitter
Instagram