[wslh]

> ARM does originate from education

But all our observations are in hindsight, who could predict that ARM/RISC would be a winner in the XXI century? Apple and mobile devices drove that wave, and Intel failing miserably in the space.

[t43562]

I think the people that created it knew it was important because prioritising performance per watt was a strange strategy and yet it was a killer strategy. ARM was in phones and PDAs long before Apple used one in an iPhone and it got there because of performance/watt. It had already reached 100s of millions of users before the iPhone was released.

The PC industry and phone industry were very different and the PC guys were sleeping at the point where ARM showed it was going to change everything. I think the most humorous thing is that the turning point was kicked off by Intel with it's StrongARM which showed everyone that ARM was much more than a microcontroller architecture. Whoever decided to drop that should win a buffoon of the century award.

Also ARM's licensing model hasn't made it a megacorp. I suggest (without data so be sceptical) that Qualcomm and Apple and Samsung and so on have all made much more money than ARM did and that's why they tolerated ARM and adopted it. So it was the fact that they weren't making some play for dominance that was absolutely critical - especially as a non-American company.

[jlarcombe]

StrongARM was DEC originally then developed into XScale at Intel when they acquired that division from Digital as part of a lawsuit settlement I believe.

The funny thing about the ARM performance/watts thing is that it wasn't originally a design goal except for the need to use a plastic rather than ceramic package to save money. They were aiming for 1 watt but ended up at 0.1 watt!

[t43562]

Dead right. Sorry.

[Symbiote]

Another article on the same site has a good summary of the history of ARM — there was plenty I wasn't aware of: https://www.abortretry.fail/p/mips-for-the-masses

[wslh]

> ARM was in phones and PDAs long before Apple used one in an iPhone and it got there because of performance/watt. It had already reached 100s of millions of users before the iPhone was released.

I put the iPhone and Apple as an example since they were very unique in customizing ARM for mobile devices. Sadly, we are waiting for M1, M2, M3, and M4 processors in the PC space.

Regarding a timeframe, it was around XXI century that ARM reached its users. There was a big time gap between the 70/80s microcomputers and mobile devices. Palm Pilot, for example, use the Motorola 68k. Psion started using ARM devices in the late 90s, before that they used an x86 variant. Newton was a failure, although it shows Apple mind clarity about ARM.