Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Status
The first post of this thread is a WikiPost and can be edited by anyone with the appropiate permissions. Your edits will be public.

Do you think the first benchmarks are correct?


  • Total voters
    314

Henk van Ess

macrumors demi-god
Original poster
Aug 20, 2008
314
241
Amsterdam
It's the fastest core on Geekbench, so why not in cinebench?
Geekbench doesn't stress the CPU, so basically a waste of time, it just presents big numbers to brag about. CineBench R20 will push the CPU, it also reveals weakness if any in the cooling system. As stated best to look at the application you use, benchmarks are only really useful for comparisons
 

MacD

macrumors regular
Feb 9, 2005
117
139
How much better can the score get plugged in?
Potentially a lot better, because it's not trying save battery life. Just depends on how Apple handles power usage with battery only vs plugged in, with intense work loads on the M1.
 
  • Like
Reactions: LancesUK

thingstoponder

macrumors 6502a
Oct 23, 2014
916
1,100
Is there a reason why ARM chips can't reach higher clock speeds?
Apple’s architecture is very wide, it’s what it’s designed for.
x86 running at higher GHz isn’t a good thing, it’s a necessity because their IPC isn’t nearly as good. It makes the power usage drastically higher if it takes a higher clock to get the same performance level.

That being said we haven’t seen what they have in store for the Mac Pro. They could go all in on power and crank them up.
 

IvanKaramazov

macrumors member
Jul 23, 2020
32
49
How much better can the score get plugged in?
I could be wrong and obviously have no evidence, but I'd be surprised if it runs much / any faster plugged in.

The potential clock difference is interesting though, referring to the suggestion that 11.0.1 adds ~100MHz to the clock. Worth noting that the Geekbench scores (probably from reviewers) all report 11.0.1 as the OS, while the shipping model that was just run through Cinebench reports 11.0. And indeed, the GB results show a clockspeed of 3.2, while Cinebench above is reporting 3.1.
 

Nikolaosth

macrumors newbie
Oct 7, 2020
23
17
from the computerbase.de list of r23 results it seems the ryzen mobile chips are quite above m1 in multicore scores( those that actually matter in 3d rendering when doing actual work) , both in under 15watt and 25watt limits .
4800u @25w 10.235 , 10.111
4800u @15w 9674

M1 single core is great . Multi fails sort in this rendering test it seems compared to amd . People who render their work using cinema4d on their laptop should take notice . The rest , not so much .
 
  • Like
Reactions: neinjohn

jeanlain

macrumors 68020
Mar 14, 2009
2,459
953
Geekbench doesn't stress the CPU, so basically a waste of time, it just presents big numbers to brag about. CineBench R20 will push the CPU, it also reveals weakness if any in the cooling system. As stated best to look at the application you use, benchmarks are only really useful for comparisons
There should be not throttling during a single-core run on the Macbook pro with active cooling. Heat should not be an issue. That is, unless Apple screwed with the fan curves.
For a muticore test, that's another story, but I'm referring to the single core result.
 

thingstoponder

macrumors 6502a
Oct 23, 2014
916
1,100
For reference, here's a quite comprehensive list of Cinebench R23 results: https://www.computerbase.de/2020-11/cinebench-r23-community-benchmarks/

M1 multicore seems to be in line with Ryzen 15W results, maybe 10-15% faster. Single core is more impressive with ~25% lead.

Seems plausible regarding the manufacturing processes and more realistic for general CPU performance than Apple's performance claims.
Those 15W Ryzens have 8 big cores and use up to 50 watts in actual usage.
 
  • Like
Reactions: Sanpete

thingstoponder

macrumors 6502a
Oct 23, 2014
916
1,100
Geekbench doesn't stress the CPU, so basically a waste of time, it just presents big numbers to brag about. CineBench R20 will push the CPU, it also reveals weakness if any in the cooling system. As stated best to look at the application you use, benchmarks are only really useful for comparisons
There’s nothing wrong with geekbench 5 other than not being a long test
 
  • Like
Reactions: Roode

bill-p

macrumors 68030
Jul 23, 2011
2,929
1,589
Potentially a lot better, because it's not trying save battery life. Just depends on how Apple handles power usage with battery only vs plugged in, with intense work loads on the M1.
For the record, the 16" MacBook doesn't have any limitation when on battery.

It runs just as fast (and hot, and loud) under battery as it does when plugged in. I can force the battery to die in less than an hour.

So while other laptops may go with lower performance on battery, I don't believe Apple does this with their MacBook Pro.
 
  • Like
Reactions: Roode and viss

LancesUK

macrumors newbie
Nov 3, 2018
23
27
This Ali King guy must be sweating right now. Whole tech internet all over him. People demanding another run plugged in. Guy will end up terrorized ?
 

MacD

macrumors regular
Feb 9, 2005
117
139
This Ali King guy must be sweating right now. Whole tech internet all over him. People demanding another run plugged in. Guy will end up terrorized ?
He should be. Of course, if results are the same, we won't believe him. :D
 
  • Like
Reactions: LancesUK

leman

macrumors Core
Oct 14, 2008
19,521
19,675
How impressive is this cinebench score?

In multi-core, its basically on par with an 8-core Intel CPU running at 60 watts. Single-core wise, it’s one of the fastest CPUs that money can currently buy. And it’s a 15W laptop with a single fan. Running without being connected to a charger.
 
  • Like
Reactions: deeddawg

cal6n

macrumors 68020
Jul 25, 2004
2,096
273
Gloucester, UK
And he’s on 11.0 @ 3.1GHz. The GB scores are on 11.0.1 @ 3.18GHz.

Can Apple overclock via a simple OS update?

They‘re just showing off now!
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.