346
Apple's A20 Pro shatters Geekbench 7 single-core record — 2nm chip beats desktop Intel Core i9 and AMD Ryzen 9 by up to 32%
(www.tomshardware.com)
A20 Pro vs Intel Ultra 7 Series 3
They probably optimize for the benchmark.. It's not uncommon
The benchmark itself measures burst compute.
For any sort of sustained workload it's nowhere near laptops, let alone near desktops.
The whole post, and especially the title and image, are misleading.
It generally is with Apple
I used to sell their gear, and there was always a lot of shitty comparisons and even boasts about things that existed on the competition, but they just didn't advertise it (because they had it so long)
Like the G5 Vector computing. Turns out, instructions like SSE on intel / AMD was basically the same thing. They just didn't brag about it anymore
Wow you must be a real engineer drawing from your vast decades long experience. And not someone sitting on a shitter pulling words out of your ass.
It didn't take long for more info to come out about the A20, and the suspected compromises that are being made on the iPhone: https://wccftech.com/apple-a20-solder-joint-leak-base-iphone-18-compromises/ . I also used to sell their gear for 3 years
Pretty much every company has been busted cheating on benchmarks, including Nvidia and Intel (its easy to do, just detect the starting frames, or/and the exe). And, I spent all of last week optimising code using benchmarks (and profiling)
On Desktop applications in particular, Apple's biggest failure has always been GPU. Which is why testers often test things like encoding (it has a chip designed to accelerate video encoding), CPU, and rarely games.
Like my M2 Mac Studio looked awesome in benchmarks (Toms hardware did one too). They didn't test games, because if they tested things like Raytracing, they would have discovered the M2 never had raytracing built in (and, computers which cost the same price DID at the time)
My Mac Pro looked awesome on paper too (it even needed FB ECC Dimms to work). The GPU was absolute slop compared to what PC's were shipping, which is why sites only ever seemed to test Apps...
If someone like Steve was testing this, I'd trust it immediately. But the entire premise of the article is around 1 benchmark... And for all we know, they might have been testing against absolute crap products