Moore's Law Alive and Well, We Will Exhaust the Periodic Table: Intel's Pat Gelsinger

[btarunr]

Intel CEO Pat Gelsinger in his InnovatiON keynote address exclaimed that Moore's Law is "alive and well," and that Intel is poised to be its faithful steward into this decade. Intel Foundry will "exhaust the periodic table" to find materials that advance silicon fabrication. The company plans to release four new semiconductor fabrication nodes in 5 years, and stated that the company's 18A node is close to building test chips on. The company hopes to see transistor-counts grow from nearly-100 billion transistors per package, to a trillion transistors per package in a decade. Gelsinger's statement comes hot on the heels of a statement by another tech giant—Jensen Huang of NVIDIA, who declared Moore's Law dead recently.



View at TechPowerUp Main Site

 

[ZoneDymo]

dem's fightin words

 

[ARF]

Competition will save us all.

Looking not good for nvidia, though.

 

[Daven]

"exhaust the periodic table"

As a chemistry PhD, I can definitely say Pat is a moron.

 

[AusWolf]

"F*** Jensen" - Pat Gelsinger, 2022

 

[R-T-B]

"exhaust the periodic table"

As a chemistry PhD, I can definitely say Pat is a moron.

Or using hyperbole and not marketing to chemists.

 

[R0H1T]

Maybe he's looking at the post Si era with that element called unobtanium?

 

[Wirko]

"exhaust the periodic table"

As a chemistry PhD, I can definitely say Pat is a moron.

Why? Eventually he and his team will find an isotope with a half-life of a few years that is usable as a silicon dopant.

 

[DeathtoGnomes]

Or using hyperbole and not marketing to chemists.

Dementia does this too.

Or using hyperbole and not marketing to chemists.

Dementia does this too.

[sarcasm]I thought Moore's had something to do with doubling something in a time period. IDK. [/sarcasm]

 

[noel_fs]

nvidia just wants to sell whatever they can manufacture the cheapest for best profits

 

[R0H1T]

i thought Moore's had something to do with doubling something in a time period. IDK.

Yes & Moore's "law" has been dead for a while now. I think it broke sometime during middle of last decade.

 

[DeathtoGnomes]

Yes & Moore's "law" has been dead for a while now. I think it broke sometime during middle of last decade.

edited hehe.

 

[AusWolf]

"exhaust the periodic table"

As a chemistry PhD, I can definitely say Pat is a moron.

"Exhaust the periodic table" ... trying to find a good semiconductor.

 

[mechtech]

"exhaust the periodic table"

Radio active CPUs - hot - even when they're off

 

[Aquinus]

Both Pat and Jensen are morons. Reality is somewhere in between.

 

[Selaya]

i mean, that coming from pat "our foundries suck ass and are still stuck on otherwise obsolete nodes" gelsinger's pretty rich tbh

 

[lexluthermiester]

Intel Foundry will "exhaust the periodic table" to find materials that advance silicon fabrication.

Then they are going to have to look beyond silicon. Which is likely what they're talking about. My guess is that the industry will likely settle on an Arsenide compound as the new substrate element for IC's as Tellurium is too rare and expensive to be used as a semi-conductor, unless they find a way to blend it in using small amounts of it. Perhaps a blend of Silicon, Arsenic and Tellurium? That might work. The chemical physics are right, but can anyone manufacture such a material?

"exhaust the periodic table"

As a chemistry PhD, I can definitely say Pat is a moron.

As a physicist with a minor in chemistry, I can say that Pat is a CEO paraphrasing something his engineers have stated to him for marketing effect, and VERY likely not without a good measure of accuracy. Your comment that he is a "moron" is itself incredibly ignorant.

 

[Count von Schwalbe]

Then they are going to have to look beyond silicon. Which is likely what they're talking about. My guess is that the industry will likely settle on an Arsenide compound as the new substrate element for IC's as Tellurium is too rare and expensive to be used as a semi-conductor, unless they find a way to blend it in using small amounts of it. Perhaps a blend of Silicon, Arsenic and Tellurium? That might work. The chemical physics are right, but can anyone manufacture such a material?

Arsenic? That takes e-waste to a new level. Pass the gloves... (I'm kidding, ok?)

Actually, he says "to advance silicon production." Mebbe he is thinking of blending, or just some other chemicals in the process. Especially if Neon becomes scarce due to Russia.

As a physicist with a minor in chemistry, I can say that Pat is a CEO paraphrasing something his engineers have stated to him for marketing effect, and VERY likely not without a good measure of accuracy.

Maybe, but his phrasing is terrible.

Your comment that he is a "moron" is itself incredibly ignorant.

Hmmm. Just decouple the statements and they can be true still though.

Actually, he strikes me as being the type to be a successful politician. Not in a good way.

 

[lexluthermiester]

Actually, he says "to advance silicon production." Mebbe he is thinking of blending, or just some other chemicals in the process. Especially if Neon becomes scarce due to Russia.

Whatever they're working on or planning, Silicon will play a smaller role in the process as IC fabrication is reaching the limits of what can be done with that element.

Maybe, but his phrasing is terrible.

Not really, it's just aimed at the general consumer. And in that context it's not inaccurate.

 

[AlwaysHope]

When they get to 1 or 2nm, what happens in the next generation?

 

[enzolt]

Then they are going to have to look beyond silicon. Which is likely what they're talking about. My guess is that the industry will likely settle on an Arsenide compound as the new substrate element for IC's as Tellurium is too rare and expensive to be used as a semi-conductor, unless they find a way to blend it in using small amounts of it. Perhaps a blend of Silicon, Arsenic and Tellurium? That might work. The chemical physics are right, but can anyone manufacture such a material?


As a physicist with a minor in chemistry, I can say that Pat is a CEO paraphrasing something his engineers have stated to him for marketing effect, and VERY likely not without a good measure of accuracy. Your comment that he is a "moron" is itself incredibly ignorant.

I mean if chemist is the furthest they got in life....

 

[lexluthermiester]

I mean if chemist is the furthest they got in life....

And?

When they get to 1 or 2nm, what happens in the next generation?

It's hard to tell if 1nm is even viable as we're talking about parts that would be near single atom sizes. For example, and depending on where you look, a hydrogen atom is around 1/10th of 1nm in size. Silicon is 2/10th of 1nm. Hafnium is a bit bigger and so on..

 

[AlwaysHope]

...
It hard to tell of 1nm is even viable as we're talking about parts that would be near single atom sizes. For example, and depending on where you look a hydrogen atom is around 1/10th of 1nm in size. Silicon is 2/10th of 1nm. Hafnium is a bit bigger and so on..

TSMC & IBM are already with 2nm, but no large volume production until 2024 & 2025. Considering the aspects of 1nm, they'll probably skip it but then what happens in the next generation?

 

[R0H1T]

2nm or 1nm is just marketing speak, they can survive on "0.1 nm" probably by the end of the decade, the real issue will be energy density & heat output. As you can already see with zen4, ADL & then likely RPL dissipating heat with such massive clocks will be super challenging on smaller nodes. Maybe TSMC will pull off their micro fin(heatsink?) designs & all will be well, but we're still some ways off of that.

 

[Valantar]

"exhaust the periodic table"

As a chemistry PhD, I can definitely say Pat is a moron.

Or just quite visibly desperate. (Not that those two are opposites: desperation can make a moron out of the best of us!) If you're on the one hand saying "this thing is alive and well, the rumors of its death are greatly exaggerated", and on the other hand saying "we will exhaust the periodic table to keep it alive", then ... well, to me that sounds like your life support systems are already failing and you're rather desperately looking for ways of improving them. Those two statements are quite explicitly contradictory.