Chester (lamchester.delete@this.gmail.com) on August 14, 2020 11:12 am wrote:
> > Do you believe that AMD can ship a substantial boost over Zen 3 by end of 2021?
> > Not to knock AMD, but that seems, uh, ambitious. Even when Intel was
> > firing on all cylinders, the best they could do was Tick Tock.
> > I'd expect Zen 3 (7nm, IPC boost) to have minor tweaks for
> > 5nm, the easy stuff (eg larger LLC) but no serious
> > IPC boost. MAYBE they can pick up a few percent in frequency,
> > but more likely, I'd expect, is they take the process
> > win as lower power/able to sustain today's maximum frequency longer/over more cores simultaneously.
>
> And with 5nm comes more transistor budget for bigger buffers, more execution resources, and wider stages.
> I think it can be a substantial boost over Zen 3. Zen 2 was a substantial boost over Zen 1 after all.
>
> > Sure, I see Apple as being able to do this and not AMD. Am I being hypocritical? No, I'm looking
> > at the track record in both spaces. Like I said, even Intel at its best couldn't do what you're
> > suggesting --
>
> AMD has a much longer history of designing CPUs than Apple, and much more eperience targeting high performance
> segments as well. They have one-upped Intel before, and are doing so now. I wouldn't underestimate them.
>
> > something about the combination of x86 and/or reaching for frequency (lotsa
> > custom circuits, trying to tweak *any* feature impacts clock)
>
> That's a very cynical view. Yes, AMD and Intel hit clock speeds higher than ARM chips, but they're still striking
> a balance. Sometimes it's done well (chips today), and sometimes it's done badly (P4, bulldozer).
>
> You seem to assume companies other than Apple are prone to making extreme,
> black-and-white design choices. That's not what happens at all.
>
> > and/or trying to target a single
> > design from server to mobile seems to make that sort of pace impossible.
>
> Again both AMD and Intel have made big gains. Sure, percentage wise they gain less than ARM,
> but that's because ARM's starting from behind (and benefit from seeing what worked for AMD/Intel).
> RISC-V starts from even further behind and thus sees even bigger percentage gains.
>
> But I do agree that trying to target a single design for everything from server
> to mobile creates compromises. That's likely holding AMD/Intel back.
>

It's also plausible that Apple can simply outspend Intel (and definitely AMD), but Maynard probably goes a bit farther with the triumphalism, beyond just this objective money factor :)

BTW on the topic of Zen 3 follow-up (Zen 4), that's going to be 5nm so it should be a big efficiency boost again, like shown by Rome, but I'd expect it in H2 2022, not in 2021.