Apple M1 Macs on the test bench

[HEXUS]

Reviews put the first Apple Silicon-powered machines and the M1 SoC into focus.

Read more.

[gagaga]

The M1 is impressive, but a lot of the press seems desperate to make it look amazing. Ars using a laptop Ryzen with Zen 2 against a desktop M1 when Zen 3 laptops are widely avaialble (still on 7nm process vs the 5nm M1).
The margin looks to be ~ a generation of Ryzen, so is likely to disappear when Zen 4 appears on 5nm (of course, Apple have updates coming next year too).

[3dcandy]

The M1 is impressive, but a lot of the press seems desperate to make it look amazing. Ars using a laptop Ryzen with Zen 2 against a desktop M1 when Zen 3 laptops are widely avaialble (still on 7nm process vs the 5nm M1).
The margin looks to be ~ a generation of Ryzen, so is likely to disappear when Zen 4 appears on 5nm (of course, Apple have updates coming next year too).

Whilst I agree it's far more worrying to me that they are making impressive claims back to Intel based laptops where Intel has not helped Apple really over the last 3 or 4 years and Apple basically called them out on it regularly

[LSG501]

Honestly I don't really trust benchmarks 100% between platforms, even cinebench which is fairly accurate, but while I'm not doubting the m1 is a fast chip (arm isn't exactly a slouch) there is a side of me which wonders if there is maybe some 'under the hood' tweaks that apple does when it detects benchmarks etc, it wouldn't be the first company to do it (they have in the past and it was pretty common on on android) and they'll obviously want to make out the decision is right etc and it's not like apple doesn't control the entire hardware/software etc.

Having said that unix is far more 'friendly' when it comes to different core architectures than say windows.

Also I wouldn't trust the zip compression test 100% either, for one thing we don't know if the storage is equal and that can have a bigger impact than more cores, not to mention there is no mention of how many cores it's even using, I'm going to guess it's just one core.

Graphics benchmarks are great and all that but are they visually the same, is 'high setting' on mac the same quality as 'high setting' on windows using dx12 etc... just because something says it's 'high setting' doesn't mean the quality is the same across platforms.

[kompukare]

Ars using a laptop Ryzen with Zen 2 against a desktop M1 when Zen 3 laptops are widely avaialble (still on 7nm process vs the 5nm M1).

While you were in your time machine, were you able to bring back next weeks Euro Millions lottery numbers?

(That is, Zen 3 has just launched on desktop, Zen 3 APUs are at a few more months away with hopefully a CES announcement in January.)

As for M1, well the GPU is actually very impressive in that one tomb raider game they benched especially since the binaries were still x86. GTX 1650 in a 20-25W SOC is pretty impressive.

[gagaga]

While you were in your time machine, were you able to bring back next weeks Euro Millions lottery numbers?

(That is, Zen 3 has just launched on desktop, Zen 3 APUs are at a few more months away with hopefully a CES announcement in January.)

As for M1, well the GPU is actually very impressive in that one tomb raider game they benched especially since the binaries were still x86. GTX 1650 in a 20-25W SOC is pretty impressive.

My bad - confused the 5000 line with the 4000.

[ritiklahori270]

seems good