By: Groo (charlie.delete@this.semiaccurate.com), October 30, 2016 9:28 pm
Room: Moderated Discussions
Doug S (foo.delete@this.bar.bar) on October 25, 2016 8:59 am wrote:
> Great article, David! I had been wondering why Apple had been hiring GPU engineers for
> years but weren't doing a custom GPU. Turned out they were, we just didn't know it.
>
> Interesting that they seem to be taking a slower approach towards a full custom GPU than the
> 'big bang' they did with the A6 on the CPU side. Do you think this might have something to do
> with patents? i.e., so long as they are co-designing with Imagination, they can effectively
> lean on Imagination's patent agreements, whereas if they cut them out and went full custom they'd
> need a large GPU patent warchest to obtain the necessary cross licensing agreements.
Err, I wrote it up with deliverables dates, which SoC it would intercept, and a lot more. No links due to RWT policy but I wrote it up ~2 years ago.
-Charlie
> Great article, David! I had been wondering why Apple had been hiring GPU engineers for
> years but weren't doing a custom GPU. Turned out they were, we just didn't know it.
>
> Interesting that they seem to be taking a slower approach towards a full custom GPU than the
> 'big bang' they did with the A6 on the CPU side. Do you think this might have something to do
> with patents? i.e., so long as they are co-designing with Imagination, they can effectively
> lean on Imagination's patent agreements, whereas if they cut them out and went full custom they'd
> need a large GPU patent warchest to obtain the necessary cross licensing agreements.
Err, I wrote it up with deliverables dates, which SoC it would intercept, and a lot more. No links due to RWT policy but I wrote it up ~2 years ago.
-Charlie