• Welcome to ASR. 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

EERecordist said:
Itanium was driven and paid for by HP. It was successful in its product cycle. Intel's weakness is its engineers successfully lobbying for overly complex implementations of the X86 architecture over market needs. That killed Larrabee which might have allowed them to catch up in graphics, then AI. They had the information to succeed in AI and missed that. ARM, and mobile were strategic mistakes. Whoever screamed the loudest won.
Click to expand...

Well, I covered Intel as a journalist during that period and talked to all the principals mentioned and more. We are both broadly correct. I would quibble with your "HP in the driver's seat" characterization. Intel was nobody's, er, batch. Or as it was put to us in engineering school when learning to read resistor codes, they gave willingly.

The point I originally made, to which you were replying was: "Craig Barrett is one of the people responsible for where Intel is today. He championed Itanic (Itanium) and when he was chairman, he and his incompetent CEO Paul Otellini (who sold off their mobile ARM IP) started Intel on its sad, sagging path." I stand by this 100%.

In your reply, where you talk about "complex implementations of x86" being elevated over market needs, I'd agree. That's why AMD was able to get the drop on Intel, beginning with AMD's Opteron, which debuted AMD's Hypertransport architecture in 2003 (and then in multicore in 2005), eliminated the Front Side Bus.

But I would add a little specificity to point that Intel prioritized what they wanted over market needs by noting that the driver for them not doing an integrated bus architecture was because they wanted to keep selling separate core logic chips as opposed to integrating more functionality on the processor die/package. One of Intel's key chip engineers went public with this many years ago; I can't remember the reference.

There was another factor, too, but I can't remember it right now. (I'm old.) When it pops back into my memory, I'll edit this reply.
 
If Intel really stops making chips, it will be the end of an era. I cannot fathom this. I like TSMC, but do we really want to rely on them to be the only high-end chip manufacturer on the planet? Especially with the China-Taiwan situation? We need competition.
 
digitalfrost said:
If Intel really stops making chips, it will be the end of an era. I cannot fathom this. I like TSMC, but do we really want to rely on them to be the only high-end chip manufacturer on the planet? Especially with the China-Taiwan situation? We need competition.
Click to expand...
Would you settle for Tesla?

I suspect that the future belongs to dedicated AI chips, and power efficiency will be the competitive parameter.
 
Tesla is not manufacturing chips, in fact they just made a deal with Samsung to make their chips for them. I am not talking about chip design. The death of x86 has been called a long time, and still we use these CPUs everyday. I assume this will continue, with added AI capabilities, either on the GPU or with dedicated AI parts .
 
Intel's foundry includes packaging. It competes with TSMC's advanced packaging like https://3dfabric.tsmc.com/english/dedicatedFoundry/technology/3DFabric.htm

It may be found in mobile phones, laptops, desktops, and servers. Probably home audio does not need it, yet!

www.wired.com

The Ridiculously Nerdy Intel Bet That Could Rake in Billions

Advanced chip packaging is suddenly at the center of the AI boom. Intel is going all in.
www.wired.com www.wired.com

www.intel.com

Advanced Packaging Innovations | Chip Packages

See how Intel is enabling tomorrow's semiconductor chip packaging to deliver a systems foundry for the AI era.
www.intel.com www.intel.com
 
Last edited:
Tom C said:
Things don’t sound promising for Intel.
Click to expand...
I'm not expert in the field but I do have a warm spot for Intel.
As somewhat of computer nerd for the last nearly 30 years, I watched them move the SOTA in processing forward miles beyond what I ever would have imagined in 1998. It took a huge dedication to the field in money and development to get where we are today. Times change along with corporate market positions but I do still wish them well in future days.
 
You must log in or register to reply here.
Back
Top Bottom