What Intel should have done was gone into the cloud services ... remember we used to have internet services 20 years ago .. We used to have very advanced data centers internally, old fabs got converted to data centers. Now cloud providers like Amazon, Microsoft, and Google are designing their own CPUs
post id: @gxx+18wJnvUO What do you folks think about this? Cloud services train left the station long time ago but I wonder if there is still a chance to fully jump on board?
Huh?
Intel used to have a division called "Intel Online Services' and it was shutdown after a couple of years of operation. It was started when Intel saw the market cap for exodus. Several of my friends working there lost their jobs (after 2 years of partying) when they shut that down.
Those scripts save money on licenses for real software and have saved an accumulated negative 40 million engineer hours.
The only software required to run a cloud service is a perl script written by a manager with EE degree and no software background who uses global variables because they can't figure out how to pass variables to functions. Just throw a bunch of expensive Intel cpus at it and it will run great. You guys are just naysayers.
I would say you need to go as far as the big bang. That’s when all the fate was really decided.
Seriously guys. 50 years? Really? You work for intel?
And we wonder why Intel is doing so bad.
Actually you need to go back almost another hundred years to the invention of the light bulb and GE.
Intel's fate was sealed 50 years ago when it decided to become a parts manufacturer, instead of systems integrator, much less service provider. That position is always going to be lowest on industry totem pole, most restrictive (can't compete with customers), will be target of cost-cutting, commodification and race to bottom. The period from about 1990-2006 when Intel really called a lot of the shots in the industry was probably an aberration.
I have seen how ancient the skills are for software engineers in Intel. Many of them coming from no name local colleges without even a proper degree, much less in computer science. They were embarrassment for the title of software engineers.
A cloud business couldnt have possibly been achieved by those folks. Most of them couldn’t design a system on that scale or anything remotely close. They got hired for the lowest of salaries in terms of software engineering market. And their quality justifies it.
Intel wouldn't have succeeded in the space anyway. The biggest cloud service providers today are AWS, Microsoft, and Google. All 3 of them are known for their software prowess which is what you need to be a cloud service provider. Intel would have been willing to invest top $ to lure top software engineers and to convince them to move to low cost geo's like OR and AZ. It couldn't, or wasn't willing, to do with it hardware engineers, so I don't see why it would do it with SW engineers.
You NEVER compete with your biggest customers.
What happened years ago was to nurture new business that would sell more CPUs.
That is why foundry failed too, besides the terrible service from LTD you couldn’t trust they would run your silicon versus Intel’s own much more valuable x86.
Intel has no reason to be in silicon wafer fab without process leadership and they don’t have leadership now. Competing against AMD, Apple, Nvidia and others who use TSMC or Samsung. I guess they can be like microchip and On and accept a far lower margin soon as the leaders gobble up the leading edge silicon