News Intel completes assembly of first commercial High-NA EUV chipmaking tool as it preps for 14A process development in 2025

[Admin]

Intel Foundry announced it had completed the assembly of the industry's first commercial High Numerical Aperture (High-NA) Extreme Ultraviolet (EUV) machine in its D1X fab in Oregon.

Intel completes assembly of first commercial High-NA EUV chipmaking tool as it preps for 14A process development in 2025 : Read more

 

[JayNor]

"TSMC then armed Intel's rivals, like AMD, with its better process tech, leading to significant market share losses for the product side of Intel's business."

So Intel's introduction of Sierra Forest and Granite Rapids chips on Intel-3 should be expected to regain its process tech lead and recover significant market share over AMD?

 

[Giroro]

Intel 14A is not 1.4nm The name has nothing to do with the size. They were very clear on that when they when they tried (and succeeded) to get away with their misleading rebrand of 10 nm to Intel 7.

They never gave us a roadmap of what process got renamed to 14A, but it's likely 2 or 3 nm. Granted transistor size has always been a big marketing trick; when they named their transistors in "nm", it still had little to no relation to physical size of any feature on any chip. But the point is 14A is almost certainly not going to outperform whatever tech TSMC is calling 2nm.

 

[peachpuff]

Intel says it is now on a course to retake the lead by launching five nodes in four years

 

[thestryker]

After seeing how much bigger the High-NA machines are I couldn't help but wonder if that was going to impact adoption for some fabs. Intel mentioned that these cannot just be swapped in where EUV machines were unless the fab was already designed for it. It will be interesting to see how quickly ASML can ramp up production since the current rate is fairly low.

A couple of article mistakes:

Intel finally adopted EUV technology with its 'Intel 7' process node, but the years of delays left it lagging several nodes behind TSMC.

Intel 4 not Intel 7

TSMC achieved First Light with its High-NA machine in February, and Intel will soon follow.

ASML not TSMC

Intel 14A is not 1.4nm The name has nothing to do with the size. They were very clear on that when they when they tried (and succeeded) to get away with their misleading rebrand of 10 nm to Intel 7.

Nothing since the shift to FinFET has been an accurate representation of size. Intel shifted the name of "10nm" because their competition is even less honest about node sizes and Intel's process is largely equivalent to TSMC N7.

So Intel's introduction of Sierra Forest and Granite Rapids chips on Intel-3 should be expected to regain its process tech lead and recover significant market share over AMD?

According to Intel's slides from their foundry presentation (Ian Cutress has them in his writeup) Intel 3 is still behind on process technology in more areas than not. 18A is when they should be hitting at worst equivalent in every area. They should be offering equivalent (or maybe even better as we haven't seen performance numbers) compute density with SRF/GNR though.

 

[Batweasel]

The jumpsuits and little red hats make this look like if SuperHot did a 70s James Bond supervillain lair level.

 

[craigss]

Research ?
On course?
Well they "researched " 10nm for many years whilst fools kept throwing money at them +++++++ many times, tbh they are doing it again aka 12xxxx to 14xxx there is no real change yet its claimed to be a new gen chip on 14....but we all know the truth