Good article. Good to see some experienced and strong technical individuals taking the lead on their comeback. Interesting about deciding to switch to some external tool flows. It can sometimes be hard to break a “not invented here” mindset.
Not sure if you have control over the web design, but the CHIPS overlay on the front page photo completely obscures dr. Kelleher's face. Just seemed a bit off.
Maybe Ryan - the Editor-in-Chief has never gone onto Anandtech homepage on his phone. I'm not sure about other readers but 99.9% of the time I use my phone to read, whether it's an online article or an ebook. Most article's frontpage pictures on AT get blocked by black overlay banners and they always block the most important parts of the pictures. I really don't mind but it looks so unprofessional.
I must admit I haven't visited the mobile site on a phone with this article up. Long story short, the label there is a dynamic item with a fixed size, so its placement varies with the width of the screen. It looked better in a simulated browser window.
Maybe this is answered elsewhere, but either way, I'm always confused by this: What is the relationship between Intel, IBM and ASML when talking about 2nm.
ASML simply supplies the equipment, but its up to other companies to figure out how to actually get the equipment to scale the designs down? In which case does Intel simply buy patents from IBM? Or do intel and IBM engineers work directly together and share results? Do they also work/share directly with ASML?
Why doesn't ASML already specify how to do 2nm with some reference designs? Why does IBM/Intel have to figure this out? When IBM/Intel are trying to figure out how to do 2nm, are they already using ASML equipment testing different techniques? or is it more theoretical/simulation?
The way I understand it (I don't work for either ASML, Intel or IBM) is that there definitely is a lot of Hand-Holding for the customers who buy those $ 150 millions and up machines. And that's the rough starting price for the current volume production EUV scanners, the high N.A. ones will be a lot more. When it comes to high numerical aperture EUV, that is truly "bleeding edge", and it appears that Intel will be the launch customer for ASML for those. With these types of products, it's a case of "the companies that bleed together, lead together"; at least if they get it to work. The proof is not that one can do the lithography, that's just the start; you want to see an actually functional chip coming out of the fab. Of cause Dr. Kelleher wouldn't have given information about what else is entailed when going to 2 nm and lower; I imagine just getting the etching processes alone working probably take tens of thousands of work hours, if not more.
Another great interview. I've really enjoyed seeing the insight you've gotten into Intel especially over the last year. It seems like they've got a lot of smart and enthusiastic people driving the company forward.
After having watched the interview on Dr. Cutress' YouTube channel, I can definitely state that this woman should have absolutely been promoted to CEO. She deftly side-stepped Ian's questions on nodes (she has PR skills), was intimately aquainted with the needs of the company and could be great at marketing too. The only thing I suspect she lacks is a handle on the financial aspects (for the size of Intel, not in general) -- however, the tech to tech and fab to fab must also include sufficient hands-on experience with the financial side that she could have picked it up.
Only thing I suppose was the need to put on a fresh-face for Intel, especially for the shareholders who endured a few years of setbacks.
intel's new CEO is getting raked over the coals for business decisions and poor culture changes intel has made over the years. I doubt she would take it even if it was offered lol, intel's biggest problems by far were their management's continued waste of cash, brain drain from firing essential talent, and poor internal and external culture (people really hate Intel), things mostly unrelated to the engineering and operational side.
"It’s somewhat of an understatement to say that Intel’s future roadmap on its process node development is one of the most aggressive in the history of semiconductor design. The company is promising to pump out process nodes quicker than we’ve ever seen, despite having gone through a recent development struggle."
Intel's "process nodes", designed by their marketing department.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
13 Comments
Back to Article
flgt - Friday, February 18, 2022 - link
Good article. Good to see some experienced and strong technical individuals taking the lead on their comeback. Interesting about deciding to switch to some external tool flows. It can sometimes be hard to break a “not invented here” mindset.Cullinaire - Friday, February 18, 2022 - link
Not sure if you have control over the web design, but the CHIPS overlay on the front page photo completely obscures dr. Kelleher's face. Just seemed a bit off.sonny73n - Sunday, February 20, 2022 - link
Maybe Ryan - the Editor-in-Chief has never gone onto Anandtech homepage on his phone. I'm not sure about other readers but 99.9% of the time I use my phone to read, whether it's an online article or an ebook. Most article's frontpage pictures on AT get blocked by black overlay banners and they always block the most important parts of the pictures. I really don't mind but it looks so unprofessional.Ryan Smith - Tuesday, February 22, 2022 - link
Thanks!I must admit I haven't visited the mobile site on a phone with this article up. Long story short, the label there is a dynamic item with a fixed size, so its placement varies with the width of the screen. It looked better in a simulated browser window.
Oxford Guy - Tuesday, March 1, 2022 - link
The replacement photo is much nicer. Kudos to the photographer for the composition.serpretetsky - Friday, February 18, 2022 - link
Maybe this is answered elsewhere, but either way, I'm always confused by this: What is the relationship between Intel, IBM and ASML when talking about 2nm.ASML simply supplies the equipment, but its up to other companies to figure out how to actually get the equipment to scale the designs down? In which case does Intel simply buy patents from IBM? Or do intel and IBM engineers work directly together and share results? Do they also work/share directly with ASML?
Why doesn't ASML already specify how to do 2nm with some reference designs? Why does IBM/Intel have to figure this out? When IBM/Intel are trying to figure out how to do 2nm, are they already using ASML equipment testing different techniques? or is it more theoretical/simulation?
eastcoast_pete - Friday, February 18, 2022 - link
The way I understand it (I don't work for either ASML, Intel or IBM) is that there definitely is a lot of Hand-Holding for the customers who buy those $ 150 millions and up machines. And that's the rough starting price for the current volume production EUV scanners, the high N.A. ones will be a lot more. When it comes to high numerical aperture EUV, that is truly "bleeding edge", and it appears that Intel will be the launch customer for ASML for those. With these types of products, it's a case of "the companies that bleed together, lead together"; at least if they get it to work. The proof is not that one can do the lithography, that's just the start; you want to see an actually functional chip coming out of the fab. Of cause Dr. Kelleher wouldn't have given information about what else is entailed when going to 2 nm and lower; I imagine just getting the etching processes alone working probably take tens of thousands of work hours, if not more.thestryker - Friday, February 18, 2022 - link
Another great interview. I've really enjoyed seeing the insight you've gotten into Intel especially over the last year. It seems like they've got a lot of smart and enthusiastic people driving the company forward.stanleyipkiss - Friday, February 18, 2022 - link
After having watched the interview on Dr. Cutress' YouTube channel, I can definitely state that this woman should have absolutely been promoted to CEO. She deftly side-stepped Ian's questions on nodes (she has PR skills), was intimately aquainted with the needs of the company and could be great at marketing too. The only thing I suspect she lacks is a handle on the financial aspects (for the size of Intel, not in general) -- however, the tech to tech and fab to fab must also include sufficient hands-on experience with the financial side that she could have picked it up.Only thing I suppose was the need to put on a fresh-face for Intel, especially for the shareholders who endured a few years of setbacks.
whatthe123 - Friday, February 18, 2022 - link
intel's new CEO is getting raked over the coals for business decisions and poor culture changes intel has made over the years. I doubt she would take it even if it was offered lol, intel's biggest problems by far were their management's continued waste of cash, brain drain from firing essential talent, and poor internal and external culture (people really hate Intel), things mostly unrelated to the engineering and operational side.watersb - Sunday, February 20, 2022 - link
👍DigitalFreak - Friday, February 25, 2022 - link
"It’s somewhat of an understatement to say that Intel’s future roadmap on its process node development is one of the most aggressive in the history of semiconductor design. The company is promising to pump out process nodes quicker than we’ve ever seen, despite having gone through a recent development struggle."Intel's "process nodes", designed by their marketing department.
Klunk - Saturday, February 26, 2022 - link
Very informative; interesting interview partner.Keep up the good work!