Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

MacRumors

macrumors bot
Original poster
Apr 12, 2001
67,957
38,669


Geekbench scores for the newly announced Mac mini with M2 and M2 Pro chips have revealed a significant increase in performance compared to the previous-generation Mac mini and previous M1 Pro and M1 Max devices.

m2-mac-mini-screen-feature.jpg

The scores reveal that the Mac mini with M2 Pro achieved a single-core score of 1,952 and a multi-core score of 15,013 for a configuration with 16GB of unified memory. For comparison, the previous M1 Mac mini achieved a single-core score of 1,715 and a multi-core score of 7,442.

The Geekbench scores are the first we've seen for the new M2 Pro, which is also available on updated 14-inch and 16-inch MacBook Pros. While we've yet to see Geekbench scores for the updated MacBook Pros, M2 Pro's performance with the Mac mini is unlikely to differ greatly from its performance with the new Mac notebooks.

m2-pro-geekbench.jpeg

The M1 Pro in the previous-generation 14-inch and 16-inch MacBook Pro scored a single-core score of 1,734 and a multi-core score of 10,076 compared to a single-core of 1,952 and a multi-core score of 15,013 for the M2 Pro. The M2 Pro, according to these results, also beats the M1 Max, which achieves 1,727 single-core and 12,643 multi-core scores.

Geekbench scores for the Mac mini with the M2 chip have also surfaced, revealing similar performance to the M2 MacBook Air announced in June 2022.

Users can configure the Mac mini with M2 with up to 24GB of unified memory, compared to 16GB with the previous M1 model. With the M2 Pro, Mac mini models can be configured with 32GB of unified memory. 14-inch and 16-inch MacBook Pros powered with the high-end variant of M2 Max can have up to 96GB of RAM.

Both the updated Mac mini and 14-inch and 16-inch MacBook Pro are available for pre-order on Apple's website and will begin arriving to customers on Tuesday, January 24.

Article Link: First Geekbench Scores for New Mac Mini With M2 Pro Surface, Beats M1 Max
 
Last edited:
  • Like
Reactions: Irishman
Better than expected but still behind Intel 13th gen. Apple needs to bring their A Game for M3 chips.

Edit: Adding some benchmarks. The i9 chips are not out yet but leaked and the i7 are already on par with M2 Pro 12 core.


benchmarki7.jpg


Screenshot-2023-01-20-at-4-13-10-AM.png
 
Last edited:
Comparing against the 16-inch 10-core M1 Pro: 1742, 12141

So, single-core is up 12.1% (meh), and multi-core is up 23.7%. But the clock is also up 9.4%, so at the same clock, single-core is only up 2.5% and multi-core 13.1% (in part explained because there's 8+4 cores instead of 8+2).

Not terrible for 14 months, but not fantastic either.
 
Better than expected but still behind Intel 13th gen. Apple needs to bring their A Game for M3 chips.

The thermals are likely somewhere between Raptor Lake-P and Raptor Lake-U. The top Raptor Lake-P part scores roughly the same as the M1 Pro. The top Raptor Lake-H part performs very similarly to the M2 Pro, but burns more energy.

Now, Raptor Lake-HX does handily beat the M2 Pro. But then we're talking 55-157W, which is way more.

So, in terms of performance per watt, Apple is still well ahead.
 
Well of course it would. M1 Pro and M1 Max have the same CPU capabilities technically speaking, it's their GPU performance that separates them. Geekbench primarily tests CPU. Feel free to correct me on all that.

With that being said I'm not sure where the gap between M1 Pro and M1 Max (and presumably M2 Pro and M2 Max) Geekbench results comes from. Binning? Co-processors for video decoding?
 
Comparing against the 16-inch 10-core M1 Pro: 1742, 12141

So, single-core is up 12.1% (meh), and multi-core is up 23.7%. But the clock is also up 9.4%, so at the same clock, single-core is only up 2.5% and multi-core 13.1% (in part explained because there's 8+4 cores instead of 8+2).

Not terrible for 14 months, but not fantastic either.
12% is very good. 23% is very good. Especially della already powerful and efficient chip. Who cares how many cores and frenquency speed if the result is still efficient, doable and pretty fast.
 
Comparing against the 16-inch 10-core M1 Pro: 1742, 12141

So, single-core is up 12.1% (meh), and multi-core is up 23.7%. But the clock is also up 9.4%, so at the same clock, single-core is only up 2.5% and multi-core 13.1% (in part explained because there's 8+4 cores instead of 8+2).

Not terrible for 14 months, but not fantastic either.
yeah, thats basically a M1 refresh
shoot myself in the leg by waiting for M2 and not pulling the trigger on M1 Max MBP
 
Nah it's fine. Energy is a major cost and constraint these days, even in our office!

We don't need faster chips, we need more grunt-per-watt and that's what Apple are doing.
Exactly. Many still say that we shouldn't care about energy consumption on a desktop device, but here in Europe it's starting to become more and more of a deciding factor.
 
12% is very good. 23% is very good. Especially della already powerful and efficient chip. Who cares how many cores and frenquency speed if the result is still efficient, doable and pretty fast.

I point out the clock bump because there's only so many times Apple can do that. They're not gonna run the M8 Pro at 7 GHz. Eventually, they'll probably want to improve their design enough that they're back in the 2-3 GHz range.
 
Well of course it would. M1 Pro and M1 Max have the same CPU capabilities technically speaking, it's their GPU performance that separates them. Geekbench primarily tests CPU. Feel free to correct me on all that.

With that being said I'm not sure where the gap between M1 Pro and M1 Max (and presumably M2 Pro and M2 Max) Geekbench results comes from. Binning? Co-processors for video decoding?

There's not much of a gap. The 14-inch with Pro does 1737/12037, and the one with Max does 1746/12153. That's 0.5% and 1.0%, respectively. Could easily be a measurement error.

The Max is theoretically faster because it has twice the memory bandwidth available, but in practice, that almost exclusively seems to help the GPU, not the CPU.
 


Geekbench scores for the newly announced Mac mini with M2 and M2 Pro chips have revealed a significant increase in performance compared to the previous-generation Mac mini and previous M1 Pro and M1 Max devices.

m2-mac-mini-screen-feature.jpg

The scores reveal that the Mac mini with M2 Pro achieved a single-core score of 1,952 and a multi-core score of 15,013 for a configuration with 16GB of unified memory. For comparison, the previous M1 Mac mini achieved a single-core score of 1,651 and a multi-core score of 5,181.

The Geekbench scores are the first we've seen for the new M2 Pro, which is also available on updated 14-inch and 16-inch MacBook Pros. While we've yet to see Geekbench scores for the updated MacBook Pros, M2 Pro's performance with the Mac mini is unlikely to differ greatly from its performance with the new Mac notebooks.

m2-pro-geekbench.jpeg

The M1 Pro in the previous-generation 14-inch and 16-inch MacBook Pro scored a single-core score of 1,734 and a multi-core score of 10,076 compared to a single-core of 1,952 and a multi-core score of 15,013 for the M2 Pro. The M2 Pro, according to these results, also beats the M1 Max, which achieves 1,727 single-core and 12,643 multi-core scores.

Geekbench scores for the Mac mini with the M2 chip have also surfaced, revealing similar performance to the M2 MacBook Air announced in June 2022.

Users can configure the Mac mini with M2 with up to 24GB of unified memory, compared to 16GB with the previous M1 model. With the M2 Pro, Mac mini models can be configured with 32GB of unified memory. 14-inch and 16-inch MacBook Pros powered with the high-end variant of M2 Max can have up to 96GB of RAM.

Both the updated Mac mini and 14-inch and 16-inch MacBook Pro are available for pre-order on Apple's website and will begin arriving to customers on Tuesday, January 23.

Article Link: First Geekbench Scores for New Mac Mini With M2 Pro Surfaces, Beats M1 Max
It’s Tuesday 24
 
Better than expected but still behind Intel 13th gen. Apple needs to bring their A Game for M3 chips.
trust me, the difference is not that big..you barley see it...but on the heat and watts consuming is HUGE
Intel on x86 is like aspirated engines in a world of turbo and hybrids...huge consumes and offers more pollution with every horse power added
 
So, in terms of performance per watt, Apple is still well ahead.
…and since the bulk of Apple’s Mac business is ultra-portable laptops and ultra small-form-factor desktops, that’s the key measure of their success. Also, this is just Geekbench - application benchmarks that take advantage of the media engine, neural engine etc. may show a greater advantage over Intel.
 
For the M2 pro to beat M1 Max in both SC and MC after just 1 generation, its impressive
I wonder if M2 Max can approach the level of M1 Ultra, in SC yes, but in MC?
But i still suppose the gpu will be more powerful on the M1 Max compared to M2 Pro
 
  • Like
Reactions: KPOM and Killroy
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.