Intel Considers Abandoning 18A Node for 14A Chipmaking Process

[kondamin]

Is IBM bleeding into irrelevance? When did it happen?

IBM has always worked and made billions of dollars in large-scale computing. What company do you think the US government and large companies (like banks) hire for their operations?

IBM used to be a computing innovation powerhouse, now it lives off of patents and remaining contracts with governments and banks, which are stuck with the power arc because all the software was written 60 years ago.

it's a shadow of the behemoth it used to be, a shadow with a stack of money which buys them harddrives to host databases.

 

[mikesg]

How does "the 18A process was failing to attract new customers" turn into abandoning the 18A node? They're using it through 2028...

 

[wheresmycar]

I get it BIG corps ain't our friends hence no tears are due but...

...it does make me a bit sad to see Intel struggling these days. I grew up with them, from pre-builts and laptops to nearly every PC I built. Every upgrade I got excited about, the big ole overclocking phase which was fun and challenging, years of work machines, side projects, home entertainment, business ventures - all powered by that little 'Intel Inside' sticker. Basically the backbone of my whole computing journey. Watching them fall behind after being such a massive force in the industry feels like losing touch with an old friend but hopefully one day we’ll cross paths again. lol it ain't that bad, i'm running a bunch of intel machines at work and home but the main gaming rig switched to a 5800X3D.

...and thats nothing to do with "loyalty"... just one of those strange cosmic alliances which feels a little weird to no longer be excited about.

 

[RandallFlagg]

How does "the 18A process was failing to attract new customers" turn into abandoning the 18A node? They're using it through 2028...

This may well be a fake story. The source, Reuters, has done this many times in the past. It happens to be a blackout period for Intel so nobody's talking.

 

[Ultron1337]

My opinion withstanding that Intel should spin off its fabs.

They can't. US Gov came them aid and one condition was, that Intel must keep majority stake in their foundries.

 

[Nhonho]

They can't. US Gov came them aid and one condition was, that Intel must keep majority stake in their foundries.

Prove it.

 

[Ultron1337]

Prove it.

You can't google, maybe you can read https://money.usnews.com/investing/...deal-restricts-sale-of-its-manufacturing-unit

 

[megaclite]

What a hell ?!

 

[Deorc Mearh]

I don't want to say I know the actual Intel situation or that I know exactly what it should be done and said.... but I'm wondering if this new CEO is totally demented, or he is paid for destroying Intel and lowering its price so that other companies can buy it for a cheap price.... or the tests they did with 18A process were disastrous and he has to fake a different strategy. If you spent a ton of money on a new process and it works even without any relevant incrrease in performance you MUST use it. What I've thought in the recent past is that just reduzing the size now it's not enough. TMSC adopted 2-3 schemes for packing mos reducing the space they occupy, that I 've learnt Intel didn't adopt. So it might be they've just realized the obvious: reducing from 20A to 18A without those strategies for packing mos is not enough. Anyway claiming now you wanto to trash 18A for 14A is an insane communication strategy that might throw the stocks under the ground. The old CEO after all was pointing in the right direction to me, especially developing the graphics sector.

 

[R-T-B]

IBM used to be a computing innovation powerhouse, now it lives off of patents and remaining contracts with governments and banks, which are stuck with the power arc because all the software was written 60 years ago.

it's a shadow of the behemoth it used to be, a shadow with a stack of money which buys them harddrives to host databases.

But it isn't exactly going anywhere.

For a company, that's kind of valuable.

What a hell ?!

View attachment 406459

Older article.

 

[Tomorrow]

But it isn't exactly going anywhere.

For a company, that's kind of valuable.

Not in today's age where everything has to constantly grow and beat last quarters record.
To quote a TV show: "steady isn't sexy"

 

[R-T-B]

Not in today's age where everything has to constantly grow and beat last quarters record.
To quote a TV show: "steady isn't sexy"

The philosophy is "sink or swim" but when you have government contracts mapped out decades that philosophy flees fast.

 

[redeye]

intel needs people not paper (stock buy back)… people are a drain on “money” but research tends to do that. but you only get research for *free” when it was has been “used up” as in china can use it. TSMC will not give you their research, so intel should spend money on people (education etc) not paper (stock)

yes, stock makes “ immediate money” and people “ cost the big bucks, but can also make the big bucks in the end”
or tsmc “ equipment, equipment paper”
or intel “ paper paper equipment”
where Paper = stocks.
but intel knows this, and purposely wastes money on stocks… hence, the sad situation they’ve gotten themselves in…IMO

 

[kondamin]

intel needs people not paper (stock buy back)… people are a drain on “money” but research tends to do that. but you only get research for *free” when it was has been “used up” as in china can use it. TSMC will not give you their research, so intel should spend money on people (education etc) not paper (stock)

yes, stock makes “ immediate money” and people “ cost the big bucks, but can also make the big bucks in the end”
or tsmc “ equipment, equipment paper”
or intel “ paper paper equipment”
where Paper = stocks.
but intel knows this, and purposely wastes money on stocks… hence, the sad situation they’ve gotten themselves in…IMO

Yes intel is run by pencil pushers, but they did invest in a heck of a lot of research and machines and realestate and that last one is what killed them.

 

[Scrizz]

Feels like they put someone in place to tear the company down

That's what I've felt too.
:/

 

[TheinsanegamerN]

But it isn't exactly going anywhere.

For a company, that's kind of valuable.

That's only valuable if it can be leveraged in some way.

IBM only exists for legacy use. You know what isn't being rapidly expanded? Applications written for mainframe. That's what we call being on life support. IBM is absolutely screwed if anything ever comes along to replace those dinosaur applications. Everyone says it'll never happen.....but it did, in education. No reason it can't happen in banks or the government.

Compared to other companies, IBM is a shell of its former self. Sure, the steady business looks good, so long as you discount what the company was capable of in the 70s and 80s. If that is Intel's future, then we can look forward to tens of thousands more layoffs, fab closures, a complete withdrawal from the consumer market, and near total market irrelevance.

Not a good look for them. It's like a car company's promising future being compared to AMC

Older article.

Do you have a newer one that contradicts?

 

[R-T-B]

IBM only exists for legacy

They literally build modern super computers, so not really. It's admittedly a smaller portion of their business than historically I will grant though, and not terribly competitive. Contracts maintain them.

Do you have a newer one that contradicts?

You are posting on it.

 

[RUSerious]

I disagree. Ivy Bridge wasn't a lot better than Sandy Bridge. 22 was good at Haswell refresh. That's 2 years of delay. 14 was good at Coffee Lake that's 4 years of delay. 10 was a good one only after they renamed it to 7, that's 5 years of delay. Intel 4 (initially 7nm) is now 7 years after its initial expected date and it still isn't good. So for me the last good one was 32nm. On the other hand they had problems with 90nm. Well Netburst was not a good architecture, but Prescott was partially caused by 90nm not being able to handle higher frequencies as initially expected. Back then they were able to overcome this by changing architecture. That was good enough up to 32nm.

The question was which of the past process nodes were good. Not which processors were. The fab can do a great job, but if the CPU designs blow it, the end user doesn't see advancement. That is what happened with Intel, by in large, at first with Intel. Now even the fabs are floundering. They sure are in a pickle.

 

[Visible Noise]

IBM only exists for legacy use. You know what isn't being rapidly expanded? Applications written for mainframe. That's what we call being on life support. IBM is absolutely screwed if anything ever comes along to replace those dinosaur applications.

Those dinosaur applications like the world’s first 2nm transistor manufacturing process? You know, the one Rapidus has licensed.

The same IBM that had 2,774 patents granted last year?

The same IBM that has the lead in quantum computing?

Yep, they sound screwed.

 

[R-T-B]

Those dinosaur applications like the world’s first 2nm transistor manufacturing process? You know, the one Rapidus has licensed.

Strange. I can't find it mentioned anywhere but PR pieces from 2021? I'm not trying to be contrarian by the way, what happened to it? Or is it just radio silence for now while they work on ramping up?

2 nm process - Wikipedia

en.wikipedia.org

Not listed on wikipedia...

 

[Visible Noise]

Strange. I can't find it mentioned anywhere but PR pieces from 2021? I'm not trying to be contrarian by the way, what happened to it? Or is it just radio silence for now while they work on ramping up?

2 nm process - Wikipedia

en.wikipedia.org

Not listed on wikipedia...

IBM doesn’t manufacture much anymore, they are deep into research. They design the transistor, someone else figures out how to productize it.

First 2nm chip

Rapidus is moving to pilot production. Tenstorrent (Jim Keller) is the lead customer.

Not surprised Wikipedia has no mention, I mean it’s Wikipedia.

 

[R-T-B]

Not surprised Wikipedia has no mention, I mean it’s Wikipedia.

Wikipedia is actually usually quite good for raw node info. As always the rule with it is to check the citations.

But you answered my question anyways.

 

[Visible Noise]

You know what isn't being rapidly expanded? Applications written for mainframe.

Interesting, as IBMs software business continues to grow year after year after year.

Publicly traded companies are required to report this stuff.