Originally planned to enter mass production in the second half of 2016, Intel’s 10 nm process technology is still barely used by the company today. Currently the process is used to produce just a handful of CPUs, ahead of an expected ramp to high-volume manufacturing (HVM) only later in 2019. Without a doubt, Intel suffered delays on its 10 nm process by several years, significantly impacting the company's product lineup and its business.

Now, as it turns out, Intel’s 10 nm may be a short-living node as the company’s 7 nm tech is on-track for introduction in accordance with its original schedule.

For a number of times Intel said that it set too aggressive scaling/transistor density targets for its 10 nm fabrication process, which is why its development ran into problems. Intel’s 10 nm manufacturing tech relies exclusively on deep ultraviolet lithography (DUVL) with lasers operating on a 193 nm wavelength. To enable the fine feature sizes that Intel set out to achieve on 10 nm, the process had to make heavy usage of mutli-patterning. According to Intel, a problem of the process was precisely its heavy usage of multipatterning (quad-patterning to be more exact).

By contrast, Intel’s 7 nm production tech will use extreme ultraviolet lithography (EUVL) with laser wavelength of 13.5 nm for select layers, reducing use of multipatterning for certain metal layers and therefore simplifying production and shortening cycle times. As it appears, the 7 nm fabrication process had been in development separately from the 10 nm tech and by a different team. As a result, its development is well underway and is projected to enter HVM in accordance with Intel’s unannounced roadmap, the company says.

Murthy Renduchintala, chief engineering officer and president of technology, systems architecture and client group at Intel is quoted to have said at the Nasdaq's 39th Investor Conference:

“7 nm for us is a separate team and largely a separate effort. We are quite pleased with our progress on 7 nm. In fact, very pleased with our progress on 7 nm. I think that we have taken a lot of lessons out of the 10 nm experience as we defined that and defined a different optimization point between transistor density, power and performance and schedule predictability. […] So, we are very, very focused on getting 7 nm out according to our original internal plans.”

The Intel exec reaffirmed the company plans to start HVM production of client CPUs using its 10 nm process technology in 2019, with datacenter products following on a bit later. That said, Intel is clearly not skipping any of its already announced 10 nm products, but implies that its 7 nm products may hit the market earlier than we might expect today (i.e., four years after the 10 nm).

“One thing I will say is that as you look at 7 nm, for us this is really now a point in time where we will get EUV back into the manufacturing matrix, and therefore, I think, that will give us a degree of back to the traditional Moore’s Law cadence that we were really talking about,”

“[With 7 nm] we are going back to more like a 2X scaling factor […] and then really moving forward with that goal.”

Intel has never disclosed characteristics of its 7 nm fabrication tech, but a major reduction of multi-patterning usage as well as a more traditional 2X scaling goal vs. 10 nm indicates a more extensive usage of EUVL.

According to ASML, one EUV layer requires one EUV step-and-scan system for every ~45,000 wafer starts per month. Therefore, if Intel plans to use EUVL extensively for 10 to 20 layers, it will require approximately 20 to 40 EUVL scanners for a fab with a 100,000 wafer starts per month capacity. Considering that Intel is not the only company with plans to use EUVL in the 2020s, getting the number of EUVL scanners it might need for HVM at multiple fabs may be a challenge.

Meanwhile, so far Intel has announced plans for only one 7 nm fab: the Fab 42 in Arizona. In addition, the company is going to have some 7 nm-capable capacity at its D1 facility used for development and trials (among other things).

Related Reading:

Sources: Intel, SeekingAlpha

Comments Locked


View All Comments

  • sa666666 - Thursday, December 6, 2018 - link

    Just in time before the supposed AMD announcement for Ryzen 3000 CPUs at CES in January. "Please, just wait a little longer, and we absolutely guarantee that our future product will be the best you've ever seen. Don't bother buying AMD, wait for us to catch up".

    Their marketing is showing up, just like clockwork.
  • FreckledTrout - Thursday, December 6, 2018 - link

    Intel and Nvidia have both used these tactics however they both have also delivered on the "next" thing. The best thing AMD can do is get EPYC on Zen 2 out the door soon so they have a lot of 2019 to lead in performance per watt to get as much data center adoption as possible. The cloud providers are all itching for a proper 2nd CPU manufacturer other than Intel so they have multiple channels and to allow some negotiating on prices.
  • Targon - Thursday, December 6, 2018 - link

    More like AMD will be leading in single core performance, multi-threaded performance, performance per watt, and performance per dollar(or euro).
  • moozooh - Thursday, December 6, 2018 - link

    So far there's been no indication at all that AMD is going to lead in single-core performance, or per-core performance in general. That has long been Intel's forte, which is why their Skylake-X, already a dated microarch by modern standards, can rival and outperform Ryzens having more cores.

    Leadership in Performance per watt is likely but not a given yet. Multi-threaded performance and performance/price ratio, AMD is already there.
  • 0ldman79 - Thursday, December 6, 2018 - link

    Look at the integer units compared between Zen and the Lakes.

    If AMD gets enough market share to actually get people to optimize for Zen they'll beat Intel.

    The reason it's this close is that AMD is playing Intel's game. I really want to see someone optimize for Zen. In simple programs it should be close between them, in more complex, parallel code Zen should be a good deal faster.

    Might not ever happen, it never did for the FX. There was performance left on the table with that design.
  • notb - Thursday, December 6, 2018 - link

    I'd rather see them leading in cooperation with OEMs, compatibility, distribution chain and support. They won't win a large portion of CPU market by just making fast CPUs.

    Today there's basically just one market niche where EPYC got any noticeable traction: massive datacenters with custom built systems. And even there the share is somehow tiny compared to Intel.

    EPYC has been around for a long time and I have yet to see a company buying an EPYC-powered server from the top OEMs. So it's pretty much the same story as with Ryzen PRO.
  • HStewart - Thursday, December 6, 2018 - link

    Single core - performance - seriously doubt that
    Multi-core performance - just add more cores

    It would be foolish to believe that Intel is not acting on this - with the hype of AMD along with blacklash of 10nm and security issues - Intel will likely delivery in 2019 with re-vengous with 7nm

    I think of saying from attack on Pearl Harbor "I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve" and Dune - "Sleeper has awaken" which of course is Intel - it is not overnight that a company can create a new FAB facility as indicated above in article - likely a year in development and very smart of them to keep most of it secret
  • FreckledTrout - Thursday, December 6, 2018 - link

    Agree. Intel will be back with a vengeance. I can assume Intels 7nm is going to likely be slightly better than TSMC's next gen 5nm node. I'm just hoping AMD can do very well in 2019 so they can continue the competition once Intel responds so this isn't a repeat of Athlon days. We have seen all to well what lack of competition looks like.
  • HStewart - Thursday, December 6, 2018 - link

    Just because TSMC's uses a smaller number does not mean the node is more dense.

    Keep in mind Desktop market is minimal market - unlike AMD is primary Desktop, so going to 7nm and lower power requirements will help the laptop market a lot.
  • FreckledTrout - Friday, December 7, 2018 - link

    You probably want to reread what I said as you arguing for the same point I just made. "Intels 7nm is going to likely be slightly better than TSMC's next gen 5nm node." lol

Log in

Don't have an account? Sign up now