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

Rosetta 2 Benchmarks Surface From Mac Mini With A12Z Chip

Number 41 said:
At the cost of breaking compatibility down the entirety of their product line, losing dual-boot, and hacking off everyone that has bought a machine that will now be completely useless when the transition completes in 2 years.
Click to expand...

Nah not useless. In the last transition Apple ended support for Power PC in 2009 (4 years later) and Rosetta version one lost support in July 2011 when Lion came out. The intel macs were announced in June 2005 with the same two year transition.

As Apple seems to be following the same pattern your Intel mac should still be viable in 2025 and they'll probably support and create an Intel version on the Mac Pro for a little longer.

Regarding Windows - that's up to Microsoft really, it's not really Apple's problem. I can see Microsoft wanting to have as many seats as possible so an ARM version that doesn't suck is probably 9-18 months away.

To be fair even if you purchased an ARM mac in 2021 you'll probably be thinking about upgrading in 2025 anyway when the technology matures.
 
It's going to run into performance hurdles as Microsoft SQ1 SoC on the Microsoft Surface Pro X but it's an important start. AMD/Intel x64 will remain performance kings for the time being. Been rooting for ARM on desktop but it makes more sense building performance bang for buck hackintosh with versatility of multibooting Windows and Linux.
 
  • Disagree
Reactions: hxlover904
These benchmarks are eerily close to the GeekBenches of the Qualcomm 8cx/SQ1 running on Windows 10 through Microsoft's version of Rosetta. It just reinforces that for ARM computers to perform well, they have to be running native ARM code. Thankfully, Apple has much more power to pressure developers to port their apps than Microsoft ever will. Rosetta is just a crutch to aid in the transition, hopefully at the end of this 2 year period all major apps that people use will be compiled natively for Apple Silicon and Rosetta will only be used sparingly for legacy apps.
 
falainber said:
We were told before that using GeekBench for cross platform comparisons is pointless. Could this be a contributing factor in this case?
Click to expand...

The funny thing about this specific case is that they are running Geekbench under Rosetta. So it is emulating a x86 platform at least. Which makes it semi-useful to make educated guesses about how well the developer kit compares to current Intel Macs. And it doesn't look bad for Apple at all! They are running emulated x86 code on a tablet CPU and getting scores with 70% of a MacBook Pro...
 
mannyvel said:
What this means is that if you could get rosetta onto an IPP you'd be able to get an i5 macos surface equivalent. Who's up for it?
Click to expand...
It doesn’t have to be on the ipad Pro. It can run on your mac. Take the statically-recompiled binary, and sideload that bad boy onto your ipad, and you’re good. (Other than all the App kit calls won’t work, so it’s pointless :)
 
cmaier said:
For sure the high performance cores are better when you need performance. But, for example, if all you are doing is adding a series of integers, there is likely little difference between the two types of cores. If you want to multiply 2 integers, perhaps the high performance core has a power-sucking Wallace tree that lets you multiply 2 64-bit numbers in 5 cycles, whereas the high-efficiency core uses a design with fewer reduction layers and runs things through multiple times and takes 15 cycles.
Click to expand...

Yeah, that makes sense. From what I gather the "small" cores are meant to be powerful enough to do all "off-power" tasks and then when performance is needed the other cores can kick in.

cmaier said:
Putting aside clock rates, there are lots of design choices that you can make to make a particular operation super fast or super efficient. But certain things, like addition, tend to use the same techniques either way.
Click to expand...

This is a move toward dedicated hardware acceleration. I'd expect Apple to not really try to cram out all the compute performance from their Axx-series, but rely more heavily on dedicated hardware (GPU from AMD / ASIC s.a. Afterburner) for higher-end stuff.
 
DemNoir said:
This points to rosetta 2 being a huge success. We are already getting 2020 MacBook Air performance with an iPad cpu running in emulation. Wow.
Click to expand...
If it was a simulation these would be great results. Unfortunately it's not a simulation. This was a native ARM code. The code may not be as well optimized as it potentially could but it's a native ARM code.
 
  • Disagree
  • Like
Reactions: jdb8167, tarasis, Ulfric and 2 others
MacRumors said:


While the terms and conditions for Apple's new "Developer Transition Kit" forbid developers from running benchmarks on the modified Mac mini with an A12Z chip, it appears that results are beginning to surface anyhow.

apple-developer-transition-kit-box.jpg

Image Credit: Radek Pietruszewski​

Geekbench results uploaded so far suggest that the A12Z-based Mac mini has average single-core and multi-core scores of 811 and 2,781 respectively. Keep in mind that Geekbench is running through Apple's translation layer Rosetta 2, so an impact on performance is to be expected. Apple also appears to be slightly underclocking the A12Z chip in the Mac mini to 2.4GHz versus nearly 2.5GHz in the latest iPad Pro models.

rosetta-2-benchmarks-a12z-mac-mini.jpg

It's also worth noting that Rosetta 2 appears to only use the A12Z chip's four "performance" cores and not its four "efficiency" cores.

By comparison, iPad Pro models with the A12Z chip have average single-core and multi-core scores of 1,118 and 4,625 respectively. This is native performance, of course, based on Arm architecture.


Article Link: Rosetta 2 Benchmarks Surface From Mac Mini With A12Z Chip
Click to expand...
I am pretty sure apple fan boys will buy an arm based machine as soon as it arrives. However, I really want you to use your critical thinking and let this transition be "enough" mature before you become part of this. A rule of thumb: don't buy 1st gen products. I really think apple can make this transition work. I am just saying, give apple some time before you make any purchase.
 
To anybody concerned about the results here's a perfect explanation...


MacRumors said:
Keep in mind that Geekbench is running through Apple's translation layer Rosetta 2, so an impact on performance is to be expected.
Click to expand...
leman said:
The funny thing about this specific case is that they are running Geekbench under Rosetta. So it is emulating a x86 platform at least. Which makes it semi-useful to make educated guesses about how well the developer kit compares to current Intel Macs. And it doesn't look bad for Apple at all! They are running emulated x86 code on a tablet CPU and getting scores with 70% of a MacBook Pro...
Click to expand...
 
  • Like
Reactions: ader42
PTLove said:
Honestly forget the performance, lets look at the delta between what one would expect and what we get to get a rough idea of Rosetta performance. A 30% performance penalty for Rosetta x86 -> ARM emulation is better than I would have expected. That bodes well for this transition.
Click to expand...

Maybe I am misunderstanding Rosetta 2, but why would there be much in terms of runtime performance penalty? I thought it was doing at-install translation of the x86 -> ARM calls so it then runs native from that point on. It is not doing on-the-fly emulation -- once the app runs it is going direct to ARM.
 
The results are a little bit disappointing but perhaps this should have been expected. It is disappointing in a sense that we will have to wait to see what the new Apple chips can do. For now, the claim about ARM architecture and Apple chip design team superiority remains unconfirmed.
 
  • Disagree
Reactions: tarasis
makarama said:
I am pretty sure apple fan boys will buy an arm based machine as soon as it arrives. However, I really want you to use your critical thinking and let this transition be "enough" mature before you become part of this. A rule of thumb: don't buy 1st gen products. I really think apple can make this transition work. I am just saying, give apple some time before you make any purchase.
Click to expand...

Counterpart: first generation products are often fantastic, and the opportunity cost of avoiding them exceeds the expected risk of any downsides.
 
  • Like
Reactions: ader42 and topdrawer
Unless I'm being dumb (entirely likely), but when I searched Geekbench, I saw this. Using Geekbench 4.4.1 instead of 5.2.0. Much better scores. (I admit, I skimmed the 7 pages, but didn't see it posted yet).

Screenshot 2020-06-29 at 17.11.53.png
 
What about GPU performance? As far as I know that's the biggest challenge for Apple right now. To build their own (integrated or dedicated) GPUs together with these armed processors.
 
I am not sure why performance would mean anything at this point. Isn't this just for developers to test their apps? Apple isn't going to be showing off their desktop chips at this point so they just used their iPad Pro chip instead. Seems pretty simple.
 
People seem to think that Apple plans on releasing a Mac slower then what's currently on the market with their own processors.

Why do you think Apple would do such a thing? They would look like fools and open themselves up to embarrassment.. They are out to prove a point.
 
I really wish people would understand all Apple did was throw a modified underclocked iPad Pro A12Z Arm chip going through Rosetta 2 in a Mac Mini for developers to transition... Why are people judging this like this is a new chip they built for the upcoming Arm Macs?
 
DNichter said:
I am not sure why performance would mean anything at this point. Isn't this just for developers to test their apps? Apple isn't going to be showing off their desktop chips at this point so they just used their iPad Pro chip instead. Seems pretty simple.
Click to expand...
There are some hot heads who claimed that the chips in iPad Pro are already better than the chips in Macs. Maybe that was a little bit premature on their part. The ultimate test obviously will come when ARM-based Macs are released.
 
  • Disagree
  • Like
Reactions: jdb8167 and DNichter
This is a seriously good results on 2 year hardware and using Rosetta! This benchmark is not run natively, so it's GOOD! I fail to see why people are so negative.
Secondly pretty Apple doesn't want the competition to know what they are capable as they for sure will be scared what the true performance will be and if they even have a chance to beat it.
 
Snore . . .

Wake us up when Apple ships a Mac powered by Apple Silicon, not a mashed up DTK using a chip that's 2 generations old.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back