Intel Roadmap Leak: 10nm Ice Lake in Q2, but 14nm Hangs on Through 2021

Ever since Intel announced that its 10nm process node would be delayed once more, there have been questions about exactly when the process would be introduced. Chipzilla has committed publicly to having 10nm hardware on shelves by the holiday season of 2019. A new roadmap leak would put the introduction date earlier, with limited quantities of 10nm chips available by Q2 of this year, but it also suggests that 14nm will be a large part of Intel’s product roadmap through 2020 and into 2021.

These new leaks are courtesy of Tweakers.net and are supposedly from an internal Dell presentation. As always, rumors are rumors and should be taken with a grain of salt. One point in favor of these specific rumors, however, is that they do match up with rumors we’ve heard on both sides of the 10nm question. Intel has repeatedly confirmed that it will introduce 10nm in 2019. We’ve heard rumors that those chips could show up as early as late Q2 in limited quantities, and that’s what this roadmap indicates as well. It is not clear how old these documents are, so Intel’s plans could have changed in the interim. Commercial and client roadmaps are both included, with a primary focus on laptop processors (desktop chips are referenced as well, as are entry-level Xeons and an Atom refresh).

We’ll tackle the client roadmap first:

Another point in favor of these roadmaps being accurate is that the referenced 9th Gen Coffee Lake Refresh just happened, with up to eight CPU cores, as predicted. That’ll be followed by a Comet Lake part at the tail end of the year, with up to 10 CPU cores, still built in 14nm. Ice Lake debuts in Q2 in limited quantities, but will only support 2C and 4C configurations. Ice Lake also sees action in the ultra-low-power 5W segment, with a dual-core part. A quad-core 10nm chip in the 5W bracket won’t happen until Q2 2020, with the launch of Tiger Lake. Tiger Lake is supposed to be based on Intel’s Willow Cove CPUs (the follow-up to Sunny Cove, which debuts with Ice Lake), but those chips don’t arrive for a full year and are limited to the lower power bands.

Chips like Intel’s Lakefield — that’s the combined processor with one “big core” and four Atom cores combined via Foveros — should tip up in Q2 2019. It’s not clear what products we’ll see Lakefield in; Intel has previously indicated it built this chip for a specific customer.

The Commercial Client roadmap specifically references Intel’s SIPP, or Stable Image Platform Program. This slide may not show every introduction or timeline for standard consumer parts as a result.

This slide summarizes why Intel wants a different metric for process nodes. Even compared to the "other" 14nm, its transistor density is still 1.23x higher. Note: We do not know which specific process nodes at Samsung or TSMC were compared against Intel's products.

Underlying Intel's argument for the continuation of Moore's Law is that its 10nm will be a larger improvement on its 14nm transistor density than any previous generation.

The same graph as the previous slide, but with raw numbers on transistor density instead of scaling points.

In terms of transistor gate pitch, 20nm and first-generation 14nm/16nm were roughly comparable to Intel's 22nm node. A later 14/16nm product has improved on that metric but still lags Intel's three-year-old node.

The pure-play foundries compare better against Intel in metal pitch scaling, but they still don't match Intel's 14nm, much lets its 10nm target. Again, they're projected to deploy 10nm at something closer to Intel's 14nm.

The pure-play foundry business is expected to exceed Intel's 14nm transistor density once they build 10nm products in volume, but Intel believes its 10nm density will be significantly better.

There are two interesting things on this slide. First, it confirms that Intel *will* deploy a 14nm++ node, as previously rumored, and that its first generation 10nm won't offer equal performance. Today's Kaby Lake customers will have to wait for 10nm++, expected in ~2020, to see transistor-level performance improvements above and beyond what Kaby Lake offers.

The slideshow above is from our original coverage of Intel’s 10nm disclosure and it’s the last slide that’s most important. As we wrote at the time, this slide demonstrates that Intel’s basic 10nm process node wouldn’t offer better performance than Intel’s 14nm++ product family. It hasn’t been clear whether this would still be the case when Intel finally launched 10nm. The company has had enough time to refine its original projections and node design that its fundamental characteristics could be different than what was originally projected.

These roadmap leaks suggest that whatever the differences between Intel’s original plans for the 10nm node and its current iteration, they don’t translate into better performance at high power consumption. There have also been rumors that Intel would push through 10nm and quickly target 7nm, bringing EUV online as quickly as possible. While we don’t see any 7nm chips on these roadmaps, it’s possible that Intel retains 14nm through 2021 to speed the transition from 14nm — 7nm without pausing for 10nm at most of its foundries.

Again, take rumors with a grain of salt. But if these rumors are accurate, and AMD pulls off its own 7nm transition starting off this summer, Team Green might enjoy a much longer process node advantage vis-à-vis Intel across more product families than previously anticipated.

Continue reading

SpaceX Launches ‘Better Than Nothing’ Starlink Beta

Those lucky few who have gotten invitations to try the service will have to pay a hefty up-front cost, and the speeds aren't amazing. Still, it's a new generation of satellite internet.

The Xbox Series S Is Handicapped by Its Storage Capacity

The Xbox Series S has been favorably received, for the most part, but the console's low base storage makes the Xbox Series X a better value for a lot of people.

Apple’s New M1 SoC Looks Great, Is Not Faster Than 98 Percent of PC Laptops

Apple's new M1 silicon really looks amazing, but it isn't faster than 98 percent of the PCs sold last year, despite what the company claims.

Google Kills Free Photo Storage, Changes What Counts Toward Storage Caps

Google has announced some significant changes to Photos, especially if you use the service for automatic backup.