Intel Roadmap Update: Alder Lake In H2’21, Ice Lake-SP Late This Yearby Ryan Smith on July 23, 2020 11:15 PM EST
- Posted in
- Ice Lake
- Sapphire Rapids
- Alder Lake
Among several different updates tucked into Intel’s Q2’2020 earnings report, the company included a brief update on some of their future products. While the bulk of the company’s focus is currently on their next-generation Tiger Lake CPUs, which are launching this quarter, the company is also looking at what comes after Tiger Lake, as well as the future of their highly profitable server business.
First off, Alder Lake has finally been formally outed. The successor to Tiger Lake now has an official launch window of the second-half of 2021. The 10nm chip will be for both mobile and desktops, making it the first 10nm chip that Intel has confirmed will come to desktops. Very little is otherwise officially known about the chip, but Intel’s ISA documents have previously revealed that there will be some new instructions found in that chip.
Otherwise the six-month window for kicking off production shipments is a fairly wide one for a chip that doesn’t rely on a new process node. Intel product cycles are rarely under a year long, so at first blush we’d be surprised if this was anything earlier than a late 2021 product. But with Intel’s recent 7nm delay and planned ramp-up of their 10nm process, it may be that Intel will be trying to pull it in and launch it in Q3, similar to this year’s Tiger Lake launch.
Meanwhile on the server side of matters, Intel is preparing for both Ice Lake-SP as well as its successor, Sapphire Rapids. One of the many victims of Intel’s 10nm woes, Ice Lake-SP is Intel’s first 10nm server chip. As of late the company has been riding a wave of profitability based on its server parts, so a newer part that improves on core counts and energy efficiency will be a welcomed addition to Intel’s product lineup, not to mention better able to fend off AMD’s powerful EPYC “Rome” processors.
Initial production shipments for Ice Lake-SP are set to start by the end of this year. Though Intel’s language is loose enough that this may mean that larger volumes of the chip may not ship until 2021.
Following Ice Lake-SP will be Sapphire Rapids, Intel’s second-generation 10nm server part. Along with getting Intel’s product release cadence closer to being back on track, Sapphire Rapids will play an important role in unifying Intel’s split Xeon families. Intel’s oddball 14nm Cooper Lake Xeons, which are currently shipping, support bfloat16, but Ice Lake-SP will not. For 10nm chips that support is finally being rolled into Sapphire Rapids, making the new chip the successor to both Cooper Lake and Ice Lake-SP in every way.
Sapphire Rapids will follow Ice Lake-SP by roughly a year. According to Intel’s presentation deck, chips will begin sampling in H2’2020, while CEO Bob Swan’s prepared remarks state that initial production shipments will begin at that time.
What follows these chips, in turn, will be the big question that Intel is currently wangling with in light of their 7nm delay. The company has made it clear that they intend to maintain an annual release cadence, divorced from their process roadmap if necessary. Depending on the state of their 7nm process, that may mean 7nm chips, 10nm chips, chips using dies from both processes, or even using dies from third-party fabs. Intel has opened the door to all possibilities, and their 2022 chips will likely be their first chance to embrace their new pragmatic approach.
Post Your CommentPlease log in or sign up to comment.
View All Comments
name99 - Friday, July 24, 2020 - link"since roughly the 90nm days, the process has become less and less important, and that's why Intel has been able to compete despite"
Anton, that's very much "so far so good" thinking. Intel is cruising on velocity built up five years ago, but that can't continue much longer.
- Along dimension 1 we have AMD. Still executing well, mainly limited by how many wafers TSMC can make for them. Poor AMD. But wait. isn't Huawei, a, uh, fairly large, TSMC customer going to have to stop using them soon? Seems like that's going to open up a whole lot more capacity for an enthusiastic little Intel competitor.
- Along dimension 2 we have Graviton2, and Amazon's proof, growing every day, that there's a substantial pool of data warehouse work that runs better (cheaper and/or faster) on ARM. And Graviton2 is still mainly a learning chip! Cortex X1/Graviton 3 will be out in what, a year or less...
Meanwhile for other companies envious of Amazon's situation there are Altra and TX3.
- Along dimension 3 we have Apple Silicon. And while Apple numbers are a small part of Intel's sales, Apple and AWS have given a whole new impetus to porting all sorts of infrastructure code (your MySQL, your zlib, your zend, ...) to ARMv8. And not just to port it now, but to work hard on serious optimization.
All of these are waves that have only crested in the past few months. But crested they have.
Now we wait to see the fallout over the next few years...
Samus - Saturday, July 25, 2020 - linkIn Intel's case as a volume producer, process is incredibly important because it is the decider of how much a part costs to manufacture.
This is why AMD is able to beat them on price, just like they were when their process was more advanced during the K6\Athlon era with incredibly advanced Motorola\IBM manufacturing technologies using a smaller process and gold interconnects in scalable manufacturing. They made those chips cheap and sold them cheap, really crushing Intel's credibility as a top performance chip producer during the P4 era.
Of course it would take 20 years for AMD to repeat that, particularely because of the damage done by Intel's strong-arm competitive tactics that locked them out of various OEM's.
psychobriggsy - Friday, July 24, 2020 - linkIt appears to me that marketing and MBAs have overridden any sense of risk management in the core critical process technology arm at Intel.
Risk assessments of the move to 10nm would have flagged the 2.7x density aim as very high risk. The engineers might have said 'we think it can be done', and marketing/MBAs clearly said 'do it!' but there was nobody to say 'engineers always say they can do things! we shouldn't just base the entire business' future on that'.
If Intel had aimed for 2x with 10nm it very well might have been ready a lot earlier. 7nm would probably be ready to go now.
TSMC clearly look to see what is achievable with the technology they have and the timelines they have - this is why 5nm is 1.8x. But they will be shipping millions of 5nm chips to consumers this year via Apple, Huawei, etc.
martinw - Saturday, July 25, 2020 - linkI'm happy to pile on MBAs as much as anyone, but it seems like MBAs just can't win here. Other commenters blame the MBAs for not letting engineering have their way. You are blaming them for letting engineers have their way. In the past when Intel was more engineering led, you would think the overconfidence problem would be worse, but they did fine.
The far more interesting question is what led to the overconfidence with 10nm, how did that get approved all the way up the chain, and what really went wrong on the technical side. Would love to hear from someone on the inside at some point.
sing_electric - Saturday, July 25, 2020 - link"MBA vs. engineer" is a false dichotomy; plenty of engineers get an MBA at various points in their careers. Intel's problem is that reorganizations decided that "middle management" was worthless, but those are precisely the people that are have enough knowledge and experience to ask serious questions of engineering teams while being able to translate and filter that up to the C-suite.
It's been seen time and time again that getting rid of that layer - frequently corporate veterans who've spent 10, 20+ years at the company and REALLY know its values - is great for cutting costs short term but ends up hollowing out the brain trust in the long term.
jospoortvliet - Monday, July 27, 2020 - linkYeah I dont think it is about letting the engineers do what they want or not but about having contingencies in place. They had no plan for failure and if you fail to plan you plan to fail ;-)
peevee - Friday, July 24, 2020 - linkHear hear!
MBAs would never understand the core asset of any high-tech company - its engineers, and as such wrong people and incompetent ideas bubble up to the top.
0ldman79 - Sunday, July 26, 2020 - linkMilking 14nm isn't a strategy.
10nm didn't work for years.
No one decided to milk 14nm. They tried to gain more density than the normal jump going from 22nm to 14nm and it bit them in the ass. They tried to gain more again going from 14nm to 10nm knowing how much trouble they had with the first gen 14nm.
I don't know what the issues with 7nm are but I can make a few guesses. Part of the issue is they haven't figured out how to work properly at the 10nm level so they can't apply that knowledge to 7nm. These things are done in parallel to degree but they are also depending on learning from the previous version. Can't do that if the previous version doesn't work right.
They're going to have to admit that the gains from shrinking are going to be less from here on and it's only going to get harder.
The voltages aren't dropping any more but the traces are, which is going to increase heat and resistance, which is going to roll itself into more heat and resistance...
They're going to have to look at the problem differently.
Peskarik - Friday, July 24, 2020 - linkDude, Intel was down the drain BEFORE Bob Swan joined as CEO.
SystemsBuilder - Friday, July 24, 2020 - linkTrue but he's been an executive at Intel since 2016, i.e. as EVP and such AND of course it is not just him!
As i wrote is it's him and people around him, the people who put them there and the culture they promote.
My argument was a general statement about who drives the company and that's never one person it's the board and ultimately the big investors who control the board - as I wrote above.