3 ms·
> That's not going to happen unless Apple decides to switch from ARM to RISC-V, and... why would they? They've got 15 years experience and essentially full cont
by gumby 4y ago
> That's not going to happen unless Apple decides to switch from ARM to RISC-V, and... why would they? They've got 15 years experience and essentially full control on ARM.
15? More than a quarter century. They were one of the original investors in ARM and have produced plenty of arm devices since then beyond the newton and the ipod.
I’d bet they use a bunch of risc v internally too if they just need a little cpu to manage something locally on some device and just want to avoid paying a tiny fee to ARM or just want some experience with it.
But RISC V as the main CPU? Yes, that’s a long way away, if ever. But apple is good at the long game. I wouldn’t be surprised to hear that Apple has iOS running on RISC V, but even something like the lightning-to-HDMI adapter runs IOS on ARM.
- masklinn 4y ago> 15? More than a quarter century. They were one of the original investors in ARM and have produced plenty of arm devices since then beyond the newton and the ipod. They didn't design their own chips for most of that time.
- gumby 4y agoAt the same time as the ARM investment they had a Cray for...chip design.
- masklinn 4y agoYes and? Apple invested in ARM and worked with ARM/Acorn on what would become ARM6, in the early 90s. The newton uses it (specifically the ARM610), it is a commercial failure, later models use updated ARM CPUs to which AFAIK Apple didn't contribute (DEC's StrongARM, and ARM's ARM710). <15 years pass> Apple starts working on bespoke designs again around the time they start working on the iPhone, or possibly after they realise it's succeeding. That doesn't mean they stopped using ARM in the meantime (they certainly didn't). The iPod's SoC was not even designed internally (it was contracted out to PortalPlayer, later generations were provided by Samsung). 15 times and the revolution of Jobs' return (and his immediate killing of the Newton) is a long time for an internal team of silicon designers.
- DonHopkins 4y agoBack then, a Cray-1 could execute an infinite loop in 4.7 seconds.