> To demand that Intel apologize for its integrated model is satisfying in 2018, but all too dismissive of the 35 years of success and profits that preceded it.
Many business failure stories are like that. "It worked great until it suddenly didn't."
"By 1986... Intel was already the best microprocessor design company in the world. They just needed to accept and embrace their destiny."
...they were most certainly not the best design. SPARC was just about ready to murder the VAX at this point, and more importantly, both the first MIPS and ARM samples were delivered in 1985. Either would have trounced an i386.
The iPhone was also discussed. We should remember that Intel got StrongARM from Digital, later selling the Xscale architecture to Marvel. The destructive debris of ARM within Intel was likely substantial enough to have delivered whatever Steve Jobs wanted.
However, Apple would not be happy with 32-bit ARM for long, and Intel should have realized that. If Intel had presented the equivalent of AArch64 to Apple a year before the iPhone, meeting power and code-density requirements with newly-designed 16- and 32-bit equivalents of THUMB within a 64-bit architecture, Apple might have abandoned ARM.
Intel and HP shared Itanic - had Intel and Apple shared something like the A7, history would be quite different.
The original iPhone was released in June 2007 with a 620 MHz Samsung ARM.
That was awfully close. I wonder if Apple was consulted on the sale, and what else was privately said between the two corporations.
Intel certainly had the iPhone in their grasp. The Wiki says that the Blackberry Bold used an Xscale PXA930, the Torch used a PXA940, and the Xscale IOP348 was hitting 1.2GHz. Unbelievable that Intel walked away, after pushing several mobile products with the leading smartphone manufacturer of the time.
A peer commenter mentioned IBM. As an ex-IBM'er, I certainly think that is true. The team I worked on adopted agile methodologies (which I think has been largely adopted in the emerging tech teams). But you could still see remnants of old processes and incentives, like a weird emphasis on writing patents. That was kinda shocking to me as a SE to see a patent writing as a part of my job, but I realized it was a holdover from earlier times. Ultimately, while I enjoyed my time there, broken processes for recognizing contribution was among the most important reason I jumped ship.
Useful activity for current times. Patent licensing and suits are a big money-maker and competition blocker for companies like IBM. For instance, Microsoft has made over a billion dollars on patent royalties *for Android." All that distracting paperwork paid off nicely. IBM used their legal team to crush a company selling commercial versions of Hercules emulator. Oracle vs Google. Apple vs Samsung. Intel recently threatened patent suits if anyone clones x86 to get off Intel chips.
This stuff has a real, financial benefit to these companies. So, they're justified in investing in it.
As a kid growing up with computers in the 90's, I knew IBM was doomed. Why? Because buying an IBM-brand PC when a clone cost 1/2 the price and did exactly the same stuff was ridiculous, and Microsoft owned the software you actually used (and had to make sure all your apps were compatible with). Older folks would tell me "Well, IBM makes most of its money selling to enterprises, and they can charge millions of dollars for a computer" and I'd be like "Why would you do that when you can buy a PC for a couple thousand and hire a high-school kid to program it for you?" and they'd mutter something about transaction processing and customer support and how nobody got fired for buying IBM, and I'd be like "That's stupid. Their customers are all going to go bankrupt, and then IBM will too." I was a pretty insufferable teenager, but I was right about a lot of that.
I'm 37 now, I've lived through four generations of technology (PC, web, mobile, and cloud) and retrained for each, and I find I need to make a conscious effort to try and see everything through college-student eyes. If I were a new grad just coming out of school, how would I evaluate the world? And I try to keep that in mind whenever I read eg. a thread on crypto here. The total volume of all data on the Ethereum blockchain is roughly what I would process in 10 minutes with a small MapReduce when I was working at Google. But it's very likely a college student today would be like "So? What's the point, when nobody trusts the answers you give them because you're just a giant corporation that keeps all your data secret?"
Outsiders don't tend to talk about the predictions they got wrong.
I think there's a cognitive bias somewhere about people preferring the mental model they already hold over some unknown future. Realistically, pretty much the only thing you can be certain of is that the future will not be the same as today, it's just that picking between many different futures gives you terrible odds of selecting the right one. It seems like your options are assuming that the status quo will continue, which is guaranteed wrong, vs. choosing between multiple competing alternate futures, where you are most likely wrong.
Ben doesn't offer any advice that I can see, and he makes it clear that integration has worked very well for Intel up to this point.
This seems like basic analysis, even though people seem determined to take it as advocacy.