• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

The Slow Death and Rebirth of Intel

Berwhale said:
Then perhaps we can trace it back another 4 years to 2008 when AMD divested it's Global Foundaries fabrication business to concentrate on chip design rather than the process node race with Intel and TSMC.
Click to expand...
You know to when we can trace it back? October 2014. That's when Lisa Su took the reigns.
 
EERecordist said:
I think Intel will be successful in foundry.
Click to expand...
I would not be so optimistic on Intel foundry. Chip manufacturing is not a science job (do things in a more accurate approach), instead, it's an engineering and management business (TSMC did not invent anything to produce wafers): how to deliver a new node with good yields, meet customer's product pace, make sure everything tight and cheap... etc. If one 2nm 12" wafer from TSMC costs $25000, it's difficult to image Intel could cut the price below $35000 (assume follows the EUV, fin, gaa .... schemes). => How could people accept an iphone price-tag at $3000? (one wireless telephone costs $3000 ??? really????, well I guess I am cheap.)
the business gross income of Intel was not from the foundry. foundry is a money burging thing. If you can not have 2 major clients, you are NOT good. The best hope might be building dedicated fabs for ChatGPT or Musk. (Altman said he need dedicated fabs). Of course this take at least 3+ years for a new major node. but they also said TSMC salary are not sky high ......

** TSMC 2nm wafer price : https://www.tomshardware.com/news/tsmc-expected-to-charge-25000usd-per-2nm-wafer
 
Last edited:
chnan said:
I would not be so optimistic on Intel foundry. Chip manufacturing is not a science job (do things in a more accurate approach), instead, it's an engineering and management business (TSMC did not invent anything to produce wafers): how to deliver a new node with good yields, meet customer's product pace, make sure everything tight and cheap... etc. If one 2nm 12" wafer from TSMC costs $25000, it's difficult to image Intel could cut the price below $35000 (assume follows the EUV, fin, gaa .... schemes). => How could people accept an iphone price-tag at $3000? (one wireless telephone costs $3000 ??? really????, well I guess I am cheap.)
the business gross income of Intel was not from the foundry. foundry is a money burging thing. If you can not have 2 major clients, you are NOT good. The best hope might be building dedicated fabs for ChatGPT or Musk. (Altman said he need dedicated fabs). Of course this take at least 3+ years for a new major node. but they also said TSMC salary are not sky high ......

** TSMC 2nm wafer price : https://www.tomshardware.com/news/tsmc-expected-to-charge-25000usd-per-2nm-wafer
Click to expand...
I don't disagree, but also agree that Intel might do well in fab. There are so many little things in industrial processes. Institutional memory helps and one issue with so much off-shoring in the USA companies is loss of that institutional memory in production processes. It takes some time to re-acquire that. I think fab in the USA is fully capable of becoming among the best in the world, and unfortunately you need to invest heavily and give it time to happen even after the production lines are up and running. Perhaps 5 years. Intel is still a large enough company to make that investment and allow time for it to happen.
 
Blumlein 88 said:
I don't disagree, but also agree that Intel might do well in fab. There are so many little things in industrial processes. Institutional memory helps and one issue with so much off-shoring in the USA companies is loss of that institutional memory in production processes. It takes some time to re-acquire that. I think fab in the USA is fully capable of becoming among the best in the world, and unfortunately you need to invest heavily and give it time to happen even after the production lines are up and running. Perhaps 5 years. Intel is still a large enough company to make that investment and allow time for it to happen.
Click to expand...
well, don't get me wrong. I believe Intel can succeed, but not replicating TSMC style (efficiency/budget oriented). Intel foundry was good because of Intel defined the X86 software developing environments. In my opinion, Intel should be after things TSMC can not.

For example, one crazy idea: the I/O of a computer is about 11 pin (3 for AC 110V, 8 for RJ45). If one can reduce the chip packages to similar number (11-pin IO), you can drop the working chip, of course water sealed, into a pool like a fish line. this should save 30% electricity for data center.
 
i think people here isnt intel basically a supplier of processors and maybe graphics cards?

perhaps if you are into laptops you might be interested in the Intel Centrino Evo whatever tri setup Intel is pushing (ie. cpu wireless gpu nic chipset all supplied by intel)

to normal people it doesnt matter

but to me intel right now is the provider of 'suspect cpus'... like the old FDIV Pentium thing

and so would i get a 12 13 14th edition cpu? probably not lol

but intel are 'too big to fail' - arent they 10x the size of AMD? they'll be here forever and i dont see an real decline
 
TonyJZX said:
but intel are 'too big to fail' - arent they 10x the size of AMD? they'll be here forever and i dont see an real decline
Click to expand...
People used to say the same about Kodak and Nokia, but you're probably right.
 
Some pundits would have us believe that AI is inevitable and unstoppable, but is it really either of those things?
https://www.fastcompany.com/91154806/ai-in-everything-era-pointless

I think the true state of Copilot+ PCs is "for early adopters / tech enthusiasts". It is a preview of things which might come to pass. Maybe battery life will wind up being a bigger selling point than AI.
 
digitalfrost said:
That's nice but it will not fix the CPUs which are already broken.
Click to expand...
Or any degradation that hasn't yet reached breaking point, but might cause early failure out of warranty. People are asking about that in the comments, as well as whether the RMA process will get quicker now they've accepted there's a systemic problem.
 
Just saying, but year to date Intel's stock price has declined from $47 to $31.48 while the markets have been up overall. My understanding is this reflects doubt about whether they will be able to execute their foundry plan.
 
Large numbers of 13th and 14th generation Intel desktop CPU's are crashing. Intel says a fix is on the way to prevent this, but once it happens the CPU has to be replaced.
 
Ron Texas said:
Just saying, but year to date Intel's stock price has declined from $47 to $31.48 while the markets have been up overall. My understanding is this reflects doubt about whether they will be able to execute their foundry plan.
Click to expand...

They won't. Paul Otellini set Intel on the path to where they are today by selling off ARM, not realizing that Intel could've been a player in mobile. Instead he bet oo the puny Atom chip. His predecessor Craig Barrett first set Intel on its downward glid slope with his idiotic bet on Itanium. But then Otelliini took the left seat and pointed Intel towards the mountain side.

Notably, Barrett at least was a techie (materials science, I believe). Otellini was their first non-tech type, an economics major. Decent man; horrible CEO.

It's both inexplicable and sad what's happened to what was once America's and the world's greatest semiconductor company.

A big flashing warning light in all of this for Boeing.
 
If love companies whose CEO has a tech background. Might be survivorship bias, but just look at Facebook+Amazon, they work. Lisa Su turned AMD around. As soon as the bean counters take over you need get wary. They might be able to take profits from the good base a couple of years but ultimately the companies fail.

That said, I'm slowly building a position in Intel. They are too important for the US.
 
Intel's stock has always been up and down. I bought and sold several times over the years. AMD is also always going to cycle up and down. But either/ both when they are hammered down.
 
1722549648362.png


Well...it's gonna be at least 3 years before Intel will be in a good state again. I also wonder if they really need that many people.

1722549757971.png


They are fairly valued if everything happens as expected. I would get in at 18$ an below.

1722549852467.png
 
Ron Texas said:
My understanding is this reflects doubt about whether they will be able to execute their foundry plan.
Click to expand...
foundries might be their only way out in the future, producing for others if not for themselves

IMO their biggest issue is the mentality of their old cadre still believing they are just to big to fail in any possible way which might and will come around to bite'em in the a$$, just like it did at IBM
 
Today Intel suspended its dividend and dropped about 19% in after hours trading.
 
You must log in or register to reply here.
Back
Top Bottom