Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

mattspace

macrumors 68040
Jun 5, 2013
3,344
2,975
Australia
Apple didn't buy the ARM IP. They purchased a license to use the IP.

IIRC they received a free, non-revokable licence in perpetuity, in return for IP contributions in co-founding ARM with Acorn.

Apple Silicon ARM + Apple GPU or AMD GPU, Vs. Nvidia Silicon ARM + Nvidia GPU could be an interesting next era of computing rivalry.

Where Intel and AMD go with X86-64, and Iris / AMD / Nvidia GPU is going to be an interesting parallel world.
 

deconstruct60

macrumors G5
Mar 10, 2009
12,493
4,053
IIRC they received a free, non-revokable licence in perpetuity, in return for IP contributions in co-founding ARM with Acorn.

For the original IP back then yeah. Maybe for the first couple of cycles (all of the stuff on the 3-6 year roadmap that had been laid out) . But huge give away for all possible future IP forever. Errr that would be almost criminally illegal misconduct on ARM lawyers side. Don't even know what you have ... how can you rationally price it away for some short term cash.

That old 32-bit , thumb opcode stuff is exactly what Apple has chucked into the waste can at this point. For the Newton it was useful. At this point is probably not much. Apple gots their money's worth and more.


Apple Silicon ARM + Apple GPU or AMD GPU, Vs. Nvidia Silicon ARM + Nvidia GPU could be an interesting next era of computing rivalry.

Highly likely not going to happen. Nvidia ARM + Nvidia GPU is more than likely going to be in the Server space. ( e.g., Nvidia's supercomputer product they have now which uses AMD server chips. chuck those for all Nvidia CPU+GPU+Interconnect+Network. ). Apple isn't going into that space. Even with the Mac Pro SoC. ( macOS doesn't want to chase into that Linux server dominated space. )


Samsung ARM + AMD GPU will likely land in Windows 10 space before anything Nvidia does. That would be closer to Apple space.

In Mac Pro space I doubt Apple GPU is going to be any big player other than serving as a foundation for running iOS/iPadOS apps in compatibility mode and default embedded GPU for entry Mac Pro configuration with no slot usage.
(the role where the W5500 is now. )


Where Intel and AMD go with X86-64, and Iris / AMD / Nvidia GPU is going to be an interesting parallel world.

If Intel doesn't fix their internal FUBARs quicker they will crumble at the HPC end of the scale . They are past the "hair on fire" stage. That Board of directors needs to seriously wake up... gaming the books for short term quarterly gains and stock price goosing is even more 'wrong track' if this deal goes though .
 
Last edited:

goMac

macrumors 604
Apr 15, 2004
7,663
1,694
For the original IP back then yeah. Maybe for the first couple of cycles (all of the stuff on the 3-6 year roadmap that had been laid out) . But huge give away for all possible future IP forever. Errr that would be almost criminally illegal misconduct on ARM lawyers side. Don't even know what you have ... how can you rationally price it away for some short term cash.

Future IP is something that no one knows the details of. But at the very least, ARM did give Apple a perpetual license to the ARMv8 instruction set. Apple can keep producing ARMv8 processors forever on their own without Nvidia's involvement.

And yes, Apple paid them a huge amount of money for it. But it keeps Apple completely secure from Nvidia meddling at the present. Apple could cut off all conversations with ARM and continue making ARMv8 processors forever.

ARMv9 and ARMv10 are open questions. But that starts to get really speculative.

It wasn't because of Apple's original contributions to ARM though. It was because of the dump truck loads of cash Apple sent ARM's way. Apple isn't the only one with this license either. Several other companies have these licenses, including Nvidia I believe.
 

goMac

macrumors 604
Apr 15, 2004
7,663
1,694
Nvidia ARM + Nvidia GPU is more than likely going to be in the Server space. ( e.g., Nvidia's supercomputer product they have now which uses AMD server chips. chuck those for all Nvidia CPU+GPU+Interconnect+Network. ).

The interesting thing is Nvidia has commented that they might be willing to license these designs. That means that another company could step in and build PC class components based on their server work.

Not saying it will actually happen, but it's a plausible route for Nvidia ARM chips to end up on Windows based systems even without Nvidia going after that market.

Mods: Sorry, I thought there would be an auto merge!
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.