First M2 Pro benchmarks prove big improvement over M1 Max
Geekbench results seemingly for the M2 Pro Mac mini are of course better than for the M1 version, but they also greatly exceed the M1 Max figures.
Apple has not yet shipped the Mac mini with M2 Pro, but Geekbench now includes an entry for device identified as "Mac14,12". It appears to be the new M2 Pro version of the Mac mini, in its 12-core CPU configuration, with 16GB of unified memory.
Its single-core score is 1952, and multi-core score is 15013.
Benchmark test scores may not give a great indication of how a machine will perform in real-world use, but they do give a point of comparison. Previous Geekbench scores for the M1 Mac mini, then, have scores of 1651 single-core and 5181 multi-core.
Note that this is comparing the M1 with the M2 Pro, not the base M2. But it's still a major difference. And more significantly, Geekbench scores for M1 Max were typically 1727 single-core and 12643 multi-core.
It appears that Apple has successfully increased the performance of the M2 range over the already notably fast M1. However, benchmark tests are also not definitive.
In November 2022, for instance, benchmarks for a Mac with the M2 Max leaked online and appeared to show little improvement. Then in December 2022, a separate benchmark leak showed much better performance.
The M2 Pro Mac mini is available for preorder now and will begin shipping by January 24, 2022.
Read on AppleInsider
Apple has not yet shipped the Mac mini with M2 Pro, but Geekbench now includes an entry for device identified as "Mac14,12". It appears to be the new M2 Pro version of the Mac mini, in its 12-core CPU configuration, with 16GB of unified memory.
Its single-core score is 1952, and multi-core score is 15013.
Benchmark test scores may not give a great indication of how a machine will perform in real-world use, but they do give a point of comparison. Previous Geekbench scores for the M1 Mac mini, then, have scores of 1651 single-core and 5181 multi-core.
Note that this is comparing the M1 with the M2 Pro, not the base M2. But it's still a major difference. And more significantly, Geekbench scores for M1 Max were typically 1727 single-core and 12643 multi-core.
It appears that Apple has successfully increased the performance of the M2 range over the already notably fast M1. However, benchmark tests are also not definitive.
In November 2022, for instance, benchmarks for a Mac with the M2 Max leaked online and appeared to show little improvement. Then in December 2022, a separate benchmark leak showed much better performance.
The M2 Pro Mac mini is available for preorder now and will begin shipping by January 24, 2022.
Read on AppleInsider
Comments
EDIT; n/m found them. 1869 sc and 8900 mc.
As a comparison, here's my early 2019 iMac19,1 Core i9 8c, 72GB RAM, 2TB SSD, AMD Radeon Pro Vega 48 (bought son's fully blown iMac used for animation).
Single 1309, 67% of M2 Pro mini
Multi 8021, 53% of M2 Pro mini
Compute Metal 53883
Compute OpenCL 49436
Cost (can't remember but over $5K)
M2 Pro mini, 10/16/16 32GB 2TB $2,299
Studi Display $1,599
Keyboard and Mouse $298 (keyboard with TouchID)
Total $4,096 (funny total because it's 2 to the 12th power) <80% of iMac i9
Wish I could simply plug a Mac mini into my iMac display.
The main benefits of the "Max" series are in GPU performance, memory bandwidth (double the Pro series of same generation), and the additional ProRes encoder/decoder. If you look within the same generation (i.e. M1 Pro vs M1 Max), the CPU performance gains on the Max vs the Pro of the same core count are very subtle:
- MBP 16" 2021, M1 Pro (10 core): 1742 single core test, 12141 multi-core test
- MBP 16" 2021, M1 Max (10 core): 1745 single core test, 12191 multi-core test
Source: Mac Benchmarks - Geekbench Browser
https://browser.geekbench.com/v5/compute/6253745
Dollars to donuts, the M2 Pro 8+4+19 config will have about the same GPU performance as the M1 Max 8+2+24 config in the base Mac Studio. They are basically the same GPU performance and therefore will be about the same price for the time being. I don't think you can get a M2 Max Mac Studio until June or later. A little bit better CPU performance though, but the Studio has other things it offers. Geekbench probably has a +/-15% variation in submitted test scores due to processes doing other stuff on the machine while people test. You'll need to wait awhile for hundreds of tests to get a good average.
A a desktop Skylake i9 had 125W TDP. The Vega 48 is probably somewhere around 150 W. So about 250 W total for that performance in total CPU+GPU loads. Apple just point that performance in a 150 W box, with better CPU performance, but it is probably using 70W aggregate total or so CPU+GPU combined loads, and comes with a bunch of media engines that will take the load off the GPU and CPU for a lot of stuff.
This part makes no sense though it is interesting that MacRumors copied it but then later updated the scores to be correct. The actual scores for the M1 Mac mini are: SC: 1715, MC: 7442.
Anyone awake at AppleInsider? You need to update the article with the actual GB 5 scores.
Apple designed their own driver chip to have one 5k display. But the conventions they had to use weren’t standard, and as a result, target mode had to go away. Sad, but not intentional.
ah, I didn’t see the other posts explaining this. Just to keep in mind, Apple had displays even when target mode was in effect, so having a monitor now is obviously not why it no longer works.
but the internet keyboard warriors and youtubers need content... so every cycle it's the same thing. you'd think they're engineers and computer scientists LOL
With a decent monitor; keyboard, mouse and a little extra storage the cost outlay is not that far apart, assuming you start off with a upper level Mini
Your best bet might be to sell your iMac 5K, hopefully for >$2k, and get a display of your choice plus other accessories. The OWC miniStack will go on top of a mini nice and neatly, and there are lots of mini shaped port extenders with SSD drives that can be stacked too.