A lot of people were crediting Nuvia with the SD8 Gen 3 beating the A17 on multicore and GPU and the single core gap being smaller than in a decade, but that was just ARM’s reference X4 core, not the new Nuvia stuff that as this says will come with Gen 4.
Jaguar actually had better IPC than Bulldozer which was important at the low clocks the 8th gen ran at, and gave more of their power budget to the GPUs
Bulldozer was a poor architecture choice even if it had a great node to work off. High clocks and poor IPC was not a design for the future, plus the shared FPU meant that for floating point it was like you had half the cores, they had bet all that work would move to the GPU way too early and CPU performance needs obviously never went away.
Zendozer
Pat is doing the right things for an Intel turnaround, but I hope this is just CEO talk and not actual dismissal, and this isn’t like Blackberry laughing off the first iPhone. AMD, Nvidia, Qualcomm, and Apple are all going to be throwing down in ARM by 2025, so Intel should at least have internal efforts testing and being ready to go so they don’t get caught with their pants down if it does become a major move.
You want to power off as many components as possible at idle to save power, a detachable cache is definitely a power help
Fwiw MetalFX Upscaling was released I think two WWDC’s ago for iOS and runs on everything back to A13, but for some reason Mihoyo so far has targetted A17, maybe it really needs the twice as fast neural engine to be a benefit
Of course they also do, but their gains slowing is what allowed these gaps to narrow so much, the MT lead is gone, the GPU lead is gone, and the single core gap hasn’t been smaller in many many years. And this was with them buying out the entire first run of 3nm. So it is fair to wonder if they ever make a big bound ahead again, or if we’re just at chip convergence, like the Zen moment where AMD and Intel have stuck pretty close on performance since. We’re used to Apple leading for the previous decade.
Haven’t seen it all yet, but since the G2 GPU was already going blow for blow with the A17, I assume this one is comfortably ahead. Apple’s MT and GPU leads are gone, all they retain is single core and the margin is narrowing substantially.
I wonder if they can ever separate themselves from the pack again, or this is like the Zen moment of CPU convergence where AMD and Intel have just stuck close since.
I actually see those in airports with some frequency, and I’m always struck by how much more sleek and futuristic it looked years ago than new models now. Yes that came with issues, and the thing was slow, but without keyboard jams and an M3 processor it would be very quick and ultra portable.
This is just an opinion piece with no fresh information. We’ll see when Apple’s AI efforts do come out.
They say it uses the HR sensor and some ML magic for it to be more precise now, but supported watches can use a less precise version of it in accessibility settings
At least we should have a two generation CPU leap since these are A15 cores in the M2, so it should be a bigger leap than A17 overall, and then extra thermal headroom and maybe memory bandwidth limiting how much faster RT could get etc.
I see you speak Italian
In the end the ratio that matters is power use per performance, frequency within the same architecture is generally tied to power draw but comparing different architectures can have different power/clock speed curves.