Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple Silicon M1: benchmarks and review for Mac Mini, MacBook Air and MBP 13 M1 and comparison with other brands
- Thread starter Henk van Ess
- WikiPost WikiPost
- Start date
- Sort by reaction score
- Status
Mini Mac Pro?
Not sure if anyone caught wind of this: https://www.macrumors.com/2020/11/02/apple-smaller-mac-pro-apple-silicon/ I have a decent 2019 set-up at work and would love an affordable one for home.
www.reduser.net
Scroll to Jeff Kilgroes posts. I think he postes 3 times.
Edit. You beat me
Multi-core performance seems to be similar between the M1 and the 12 core xeon trashcan. For single core, the M1 is nearly twice as fast.
You can compare the M1's to other Macs on Geekbench:
Last edited:
Most submissions in Geekbench for the 4800U are from IdeaPads that use single-channel cheap memory. No one actually cared to release a premium Ryzen laptop, so the closest you can get to a M1 Macbook is probably the Yoga Slim 7, so I think this is the most interesting comparison:
LENOVO 82A2 vs MacBookPro17,1 - Geekbench
browser.geekbench.com
It's really close but the AMD uses more power, probably around twice as much. Notebookcheck puts the battery life for surfing at 10 hours (with 61 Wh battery) vs 17 hours from 58.2 Wh for the Macbook Pro. There are of couse a lot of factors (display, Apple usually gives lower battery life than NBC and so on), but still.
Probably that's the reason Apple made claims only about the best single core low-power performance instead of going for "most powerful low-power CPU"
amd 4800u vs AS macbookpro these are the highest scores i found for both platforms . It will be interesting running rendering tasks in blender etc . Only sustained loads will show the true self of the new chip . Big enough projects in logic , longer than a few minutes renders on finalcut , performance on davinci resolve are points of interest . burst performance does not cut it for such workloads . For example 50 tracks on logic with 100 plugins and a few softsynths running make the single core advantage meaningless since more than 1 core is hit , the frequency drops and the performance is quite close to this amd part. For video/3d Nobody renders anything on a single core . Burst loads such as web browsing , previewing a pdf , opening a mail , such actions will maybe see a benefit from the monster single core .
If you look at my graphs, you’ll see two density spikes for the Ryzen, I suspect these are the two classes of laptops you speak about.
If I am not mistaken, the CPU in that particular laptop uses the 25W configuration. And of course, it’s 8 main cores vs 4.
You have to keep in mind that the AMD CPU has significantly more thermal headroom... and it has twice as many main cores. Its not that M1 throttles and drops its performance to AMD levels, but it’s that AMD has 8 cores vs. 4 in M1 to even out the multicore score. I don’t know how much the efficiency cores in M1 contribute, probably less than 10% (if I remember correctly, Anandtech measured efficiency cores on A13 being 1/3-1/4 of the performance cores)
From what I’ve gathered, it’s in the NDA that running benchmarks is not allowed on early units.
I'm not sure, the Slim 7 only accounts for about 15% of the results (and some of the scores are below 4000 for some reason, removing them makes it even less) and the second spike is much bigger. They are actually a lot of configurations so it's difficult to get a clear picture, but there are enough 7000+ scores is order to assume that this is a realistic score with a good configuration and without Windows doing something in the background.
Yes, but Intel destroyed my trust in TDPs, so I hope someone measures their power usage on the same workload.
A lot of those people in the two promo video insets Apple showed on Tuesday were already using the new Air, Pro, and Mini, so they had access prior to the announcement. Plus, Henk posted a review from someone who had one of these new Macs prior to the announcement, so we know they have been out in the wild for at least a few weeks now.
I’m seeing debates left and right about whether M1 actually beats the highest-end i9’s or not, with some arguing that it’s unimpressive that it doesn’t. Seems a lot of people aren’t getting that this is the i3 replacement.
Nice ggplots.
Did you make a script that downloads .gb5 results or did you compile them manually?
Nice ggplots.
Did you make a script that downloads .gb5 results or did you compile them manually?
The script is here if you are interested: http://pastie.org/p/44rTtSfjQ95NrRhWZhl4iL
Sorry for the terrible quality, it was literally something I put together over a coffee break, so it's ugly and most likely buggy. Number of parallel fetches is limited since GB5 web server starts rejecting your requests if you push it too far.
The quality is probably better than what I coud achieve.
I'm not familiar with some of the packages and fonctions used (nor with the tidyverse syntax), but this script doesn't use .gb5 detailed results, does it?
Great job digging that deep. I think this is almost worthy of a front page article in itself. Many were wondering what the benefit would be of fans in the 13" MBP and Mini, with some suggesting the chips were clocked higher, but you've seemed to disprove that. As you said, it's the *exact* same chip in everything, just with the fans allowing probably longer sustained performance.
This means that literally the only variation among the entire M1 series of devices is the amount of integrated DRAM. Apple is offering a "7 core" option for the Air, but people were quick to call that out as just a binning process to get use out of chips that ended up with 1 failed core. They all still started life as literally the same exact 8 core chip being produced en masse at the fab.
We should see next week what the feedback is once people get these into their hands. A lot of people seem to have shipping confirmations for a Monday delivery window.
Last edited:
thanks @Takuro .
now watch minute 11.53 of this interview with Tim Millet (Apple)
“ I think you see across the different array of machines that we announced yesterday , you gonna see M1 at its best in everyone of those enclosures , it has the ability to fit down into the MacBook Air , you put a fan on it , you put it in a bigger enclosure , like the MacBook PRO, or the MacMini , you gonna see a different level of performance and capability.”
Thanks all for adding your thoughts. I’ve included the Apple interview , the tweet with benchmarks and the comparison between the M1’s homes (mba mbp mini) in the very first posting of this thread so we have a nice round up
Added in the very first summary posting
When asked about the differentiation between the MacBook Air and the MacBook Pro, both of which are powered by the M1 processor, Apple’s Craig Federighi was “quick to point out that the latter of the two has an active cooling system“. According to The Independent, Federighi then started sketching a graph to illustrate how heat can play a role in performance.
www.independent.co.uk
When asked about the differentiation between the MacBook Air and the MacBook Pro, both of which are powered by the M1 processor, Apple’s Craig Federighi was “quick to point out that the latter of the two has an active cooling system“. According to The Independent, Federighi then started sketching a graph to illustrate how heat can play a role in performance.
Apple tells us how it made its new chip and MacBooks
Exclusive: Apple’s hardware, software and marketing chiefs explain how the company’s new processors came to exist
It's very impressive, but i3/i5/i7/i9 naming is rather arbitrary, the 4-core 10nm is an i7 and it's the closest competitor. I think the comparison should be done on performance per watt or performance per price bracket. It clearly wins the former, but the cost is unknown and it's rather difficult to estimate. Yes, you can compare laptop prices, but they are influenced by too many factors and it's rather likely that Apple has different priorities right now from recovering R&D cost which is the main expense for CPU manufacturing. They are probably way more aggressive with pricing for the M1 Air than they were with i3 Air. And 10nm Intels are way overpriced, the 8-core 7nm AMD is cheaper