[PeCaN]

Is “make a pretty generic RISC processor and license it” really “disruption”? ARM didn't do anything especially novel, they did one thing really well. They read the Berkley SPARC papers and made a simple CPU. It's the same thing as MIPS, i960, and Am29k. The difference being ARM's Acorn computer was rather unsuccessful so they licensed the ARM IP instead.

They were very much in a race with MIPS, Intel, and AMD. They won by doing the same thing everyone else did, but by doing it quite well.

Silicon Valley terminology is getting stupider by the day.

[kens]

I view the ARM as a perfect example of disruption in the Innovator's Dilemma sense. Intel has always been pushing the boundaries of processor design, using as many transistors as possible, and nobody could threaten them with better technology.

The ARM1 on the other hand was built to be simple (because the designers hadn't built a processor before). They didn't care at all about maximizing density (the layout is pretty awful). They cared about low power consumption only to the degree it meant they could use cheaper packaging.

But this low-end ARM processor managed to hit the low-power needs of mobile devices (starting with the Newton). And now the ARM processor is a serious threat to Intel, not because ARM out-raced Intel at the more-transistors game, but because ARM was a simpler, cheaper product that disrupted the market.

I think 99% of the people who use the word "disruption" haven't read The Innovator's Dilemma (and they should), but to me the ARM clearly fits the pattern in the book.

[Maken]

>They won by doing the same thing everyone else did, but by doing it quite well.

I can't but find that pretty admirable.