How about the 10nm wafers going to Xe? Aren't those significant enough, seeing they have to honor contract obligations and all?
Ponte Vicchio is using 7 nm.
How about the 10nm wafers going to Xe? Aren't those significant enough, seeing they have to honor contract obligations and all?
Wasn't there news of a certain super computer contract using Xeon and Xe on 10nm?Contract obligations to whom?
in 2021Wasn't there news of a certain super computer contract using Xeon and Xe on 10nm?
What exactly are you talking about? Can you find me those news?Wasn't there news of a certain super computer contract using Xeon and Xe on 10nm?
What exactly are you talking about? Can you find me those news?
Wasn't there news of a certain super computer contract using Xeon and Xe on 10nm?
That makes me remember that Aurora was initially supposed to come in 2018 with Knights Hill on 10nm. Doesn't Intel get punished by breaking the contract with these delays?
Wouldn't that be enough? I agree that the off-die IMC/iGPU will be a sidegrade (with the benefit of some better memory speed support, and maybe some better graphics). But I do not think the power penalty of relying on EMIB to connect the dice will be so great that a 25% reduction in power usage of the IMC and iGPU would be insufficient to offset that penalty.
As someone already mentioned in the forums, this sunken cost fallacy at it's finest. (and not something to criticize Intel about, but rather something to learn from)
Most of the 10nm Ice-Lakes I've seen in the wild are from Dell XPS lineup.
You are missing quite a bit then. The Inspiron 3000 and 5000 are both available with Icelake, and due to their price will be higher volume than the XPS lineup. Whiskey Lake laptops are plenty, but Comet Lake has much lower volume.
Best Buy also shows there are HP laptops without the fancy names("HP 14") using Icelake, and Lenovo IdeaPads with it too.
If you look at DrMrLordX's thread, you even have $300 laptops coming with the Core i3 Icelake, meaning they are likely using some Icelake chips to fulfill the demand that 14nm chips can't.
You are missing quite a bit then. The Inspiron 3000 and 5000 are both available with Icelake, and due to their price will be higher volume than the XPS lineup. Whiskey Lake laptops are plenty, but Comet Lake has much lower volume.
Wasn't there news of a certain super computer contract using Xeon and Xe on 10nm?
Off-die means it transitions between C-states slower, meaning less scenarios where it can go to idle, and the interconnect itself would increase idle power.
Again, this is not about whether 10nm is functional or not, it's about whether or not Intel's 10nm capacity & yields are able to offset the huge 14nm demand they're seeing, so huge they are publicly apologizing, so huge that Dell is forced to announce they're missing financial targets. As I said before, I never viewed the new node capacity as a necessity to fill demand for the old node (in case of semi-failure), I always considered the new node to be either the main driver of consumer demand or simply a failure that inherently lowers demand anyway. (in favor of the competition, if any)Know what they say. In theory practice and theory are the same, but in practice its not. The same reason pipe cleaner products exist. Without the learnings of 10nm being in mass production, I don't believe successors are possible. I even think its a good thing they got that single Cannonlake out so it can be characterized in practice how it does.
Depends on the idle state? If the system has entered sleep mode, you would think the interconnect could be powered down. Heck, ARM's DynamIQ already lets you power down some or all L3 cache via power gating. Pretty sure they let you power down entire core clusters and all interconnects to them as well.
This is a risk I never took into account, but top tier managers likely did and decided to go through with 10nm anyway. Maybe they did the math and considered today's loss will be offset by tomorrow's wins, maybe they just went ahead because "sunken cost".
Zen3 is going to be interesting. If the current rumors flying about are true, then it's going to be harder to procure a high end but not high end mainstream chip like a 4800X. I was going to buy a 3800X on sale today but I chose not to. If I waited 3 years to see where AMD will be, I can wait another six to seven months.Ah yes, I too look forwards to the arrival of the broadwell-tier volume meme also known as ADL-S.
But hey, they gotta show somehow that everything is fine after Zen 3's launched and Zen 4 is so soon over the horizon.
Ah yes, I too look forwards to the arrival of the broadwell-tier volume meme also known as ADL-S.
But hey, they gotta show somehow that everything is fine after Zen 3's launched and Zen 4 is so soon over the horizon.
I wouldn't read much into it other than they actually thought they would release Alder Lake-S at one point.
Nah, I'm fairly confident they'll launch something, and I'm also fairly confident it's either ADL-S or a really late Tiger Lake-S (either way it'll be mid-2021ish), but in either case it's stupid low volune and so mostly irrelevant.I wouldn't read much into it other than they actually thought they would release Alder Lake-S at one point. There's not much downside to leaving support in there in case they (yes) decide to revive it for whatever reason.
Imma add just a little bit of context here.The addition of ADL is relatively new in public drivers, and it's not a public addition because the entries are hidden.
It sounds like you can power down the off-die chips just as readily, but it never worked in practice.
It'll be there so they can claim the IPC crown on desktop and then crawl back into irrelevancy in the desktop space till late 2022, where a real comeback might take place.
Hmm, makes me wonder how viable EMIB will ever be if Intel continues having such problems. AMD doesn't even seem to try.
If Alder Lake does show up in late 2021, it won't be claiming any IPC crowns.
If Alder Lake does show up in late 2021, it won't be claiming any IPC crowns.