Friday, July 27th 2018
Intel Stuck with 14nm Processors Till Holiday 2019
Wrap your head around this: at some point in 2019, AMD will be selling 7 nm processors while Intel sells 14 nm processors. That how grim Intel's 10 nanometer silicon fabrication process development is looking. In the Q&A session of its Q2-2018 Earnings Call, Intel stated that the first products based on its 10 nm process will arrive only by Holiday 2019, making 14 nm micro-architectures hold the fort for not just the rest of 2018, but also most of 2019. In the client-segment, Intel is on the verge of launching its 9th generation Core "Whiskey Lake" processor family, its 5th micro-architecture on the 14 nm node after "Broadwell," "Skylake," "Kaby Lake," and "Coffee Lake."
It's likely that "Whiskey Lake" will take Intel into 2019 after the company establishes performance leadership over 12 nm AMD "Pinnacle Ridge" with a new round of core-count increases. Intel is also squeezing out competitiveness in its HEDT segment by launching new 20-core and 22-core LGA2066 processors; and a new platform with up to 28 cores and broader memory interface. AMD, meanwhile, hopes to have the first 7 nm EPYC processors out by late-2018. Client-segment products based on its architecture, however, will follow the roll-out of these enterprise parts. We could see a point in 2019 when AMD launches its 7 nm 3rd generation Ryzen processors in the absence of competing 10 nm Core processors from Intel. Posted below is an Intel slide from 2013, when the company was expecting 10 nm rollout by 2015. That's how much its plans have derailed.
It's likely that "Whiskey Lake" will take Intel into 2019 after the company establishes performance leadership over 12 nm AMD "Pinnacle Ridge" with a new round of core-count increases. Intel is also squeezing out competitiveness in its HEDT segment by launching new 20-core and 22-core LGA2066 processors; and a new platform with up to 28 cores and broader memory interface. AMD, meanwhile, hopes to have the first 7 nm EPYC processors out by late-2018. Client-segment products based on its architecture, however, will follow the roll-out of these enterprise parts. We could see a point in 2019 when AMD launches its 7 nm 3rd generation Ryzen processors in the absence of competing 10 nm Core processors from Intel. Posted below is an Intel slide from 2013, when the company was expecting 10 nm rollout by 2015. That's how much its plans have derailed.
72 Comments on Intel Stuck with 14nm Processors Till Holiday 2019
Think about it: if they keep refining 14nm further, it will only be harder for 10nm to surpass it later on, in speed and / or efficiency and / or power usage with the same / more cores and with / without HT, unless they come up with some kind of "new Sandy Bridge" which can give them the possibility to leapfrog the difference between a super-refined process VS an early smaller process (it's not impossible, but it's unlikely).
Intel seems to be digging a bigger and bigger hole ...
Even if Intel's 10nm isn't good enough for a full lineup one year from now, they still have one option; use 10 nm on select models, and 14 nm on the rest. They should prioritize 6-/8-core desktop and some mobile CPUs for 10nm, and with Intel's manufacturing volume they should be able to, even if the yields are so low they might have terrible margins on select products. Even if needed, Intel can take a loss on select models for a year. It's easy to forget that most of the shipped volume doesn't have to be on 10 nm immediately.
We should expect clocks to be lower on initial Ice Lake CPUs, but primarily on the boost side. Remember that Ice Lake is a major architecture unlike Zen2, so we should expect some IPC gains to offset the lower clock speed. Also remember that Zen2 will not reach the full gains of "7nm" by Q3 2019, AMD might get closer to Intel in core speed, but is extremely unlikely to surpass them.
Specifically, when the term “the holidays” is used in the U.S., it typically refers to the Christmas-New Year time period.
It's one of the reasons I hate the term.
In the performance readiness part of the pic, you can clearly see that 14nm++ is ahead of even 10nm+ in performance, albeit slightly, and the performance increase over "regular" 14nm is quite substantial. If they end up refining 14nm++ into 14nm+++, it's possible that it can be close enough to dent the 10nm++, which is why i said:Intel is in serious trouble here. On one hand, if they release 10nm as is, it will have lower clocks then the current refined 14nm and the lower yields make it costlier to produce so it's a lose - lose situation but, on the other hand, if they opt for an even more refined14nm process again, they'll be taking a serious chunk from the advantages of the newer 10nm by having this 14nm part perform as well if not slightly better, though @ a higher power envelope, so Intel will be in this exact situation later and that makes it also a lose - lose situation.
This is exacerbated by the arrival of the Zen architecture because, if AMD hadn't "rysen from the dead" (pun intended), Intel wouldn't be as pressured as it is now.
A good backup plan for 2018 would have been to backport Ice Lake to 14nm, but that's too late now.Intel claims 10nm yields is "on track" for 2H 2019.
But the yields are at least improving.But we should clearly take everything AMD says as hard truth…
One would have assumed that a company like Intel with billions to pour into R&D could overcome the issue.
If we're talking about mass production, Intel scale, that's like saying the Tesla can mass produce model X just because they did fulfil their last minute promise of weekly model 3 deliveries.This is the exact reason why I said ~And yet here we are!
The above pic is taken from this: as can be seen in the pic's subtitle, Intel's 5nm was expected to arrive in 2019 @ the time of the article (2012), with 7nm in between 10nm and 5nm.
And how is this relevant? We all know it's delayed compared to the original plans, plans which have been revised many times and are no longer relevant.
Intel have been shipping very low volumes of 10 nm since April, and since they say yields are now on track for a volume launch 2H 2019, it means they are slowly improving.
They could manage to sort out their 10nm problems even before 2H 2019 for all we know but Intel saying it won't reassure most due to the consecutive delays: if it were board manufacturers saying it, it could potentially have more credibility, @ this point.