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.
2020 $1799 MBP Unusual Benchmark Performance
- Thread starter jgorman
- Start date
- Sort by reaction score
It shouldn't, no. Tbh if it does then there was something wrong with the laptop anyway - I always stress test new bits of kit before I spend hours setting them up how I like.
How is the performance of the 13 inch running the LG UltraFine 5K? The lack of GPU power in the past has caused the previous generation 13 inch to drop frames on even light workloads. This problem has kept me using the 15/16 inch models with a dGPU. I love the portability of the 13, but don't want to give up graphics performance for basic stuff (the most I do is light 4k video editing)
With regards to the base clocks it's just what Intel will guarantee on the chips. So for instance on the i5 28W they are saying if you max all cores out, it'll never drop below 2Ghz at 28W. On the i7 they are saying it won't drop below 2.3Ghz at 28W at full load.
Now with a good cooling system and the benchmarks seen, it's going well beyond that on the i5 and the i7 is barely any better. Intel will no doubt have given themselves some room but don't forget that for instance the room temp that Apple supports goes up to 35C or 95F. I imagine no one has tested theirs in a room that hot but maybe if you were to the clock speeds of the i5 / i7 might be more down towards their base clocks. I suspect still faster but not as much.
People often forget to mention on their reviews what their ambient room temp is. You might see one review and his is running quite cool and the other quite a bit hotter but also their rooms might be 10C or more different in temp.
Interesting reading the temp limits though, it hit 40C here last year in the UK and most of us stuck working from home without AC. That would put our rooms outside of the maximum operating temp Apple supports for their products. Thankfully I got AC at home after last year because it was awful and I already worked from home before this virus.
That was me, thanks a lot for that. I watched your video yesterday and gave you a thumbs up on Youtube
I see from the screenshots the screen recording had a bit of an impact but overall was a good result. Mine has shipped at last and should have it Monday. Don't have a 4k monitor to test it with but if I have the right cables / docks needed I'll try it with my 4K TV to get an idea myself before I order some 4k monitors.
Now with a good cooling system and the benchmarks seen, it's going well beyond that on the i5 and the i7 is barely any better. Intel will no doubt have given themselves some room but don't forget that for instance the room temp that Apple supports goes up to 35C or 95F. I imagine no one has tested theirs in a room that hot but maybe if you were to the clock speeds of the i5 / i7 might be more down towards their base clocks. I suspect still faster but not as much.
People often forget to mention on their reviews what their ambient room temp is. You might see one review and his is running quite cool and the other quite a bit hotter but also their rooms might be 10C or more different in temp.
Interesting reading the temp limits though, it hit 40C here last year in the UK and most of us stuck working from home without AC. That would put our rooms outside of the maximum operating temp Apple supports for their products. Thankfully I got AC at home after last year because it was awful and I already worked from home before this virus.
That was me, thanks a lot for that. I watched your video yesterday and gave you a thumbs up on Youtube
I see from the screenshots the screen recording had a bit of an impact but overall was a good result. Mine has shipped at last and should have it Monday. Don't have a 4k monitor to test it with but if I have the right cables / docks needed I'll try it with my 4K TV to get an idea myself before I order some 4k monitors.
That's a fair point on the ambient temps - I'll mention it next time. My 2018 i9 would regularly drop below clock under load, it was really irritating. It took me taking a colleague I'd found with the same machine in to the Apple Store to convince them mine was stuffed.
It was very nearly the last Apple product I ever bought!
[automerge]1589564792[/automerge]
Surprisingly good! I was happily using it for FCPX on 4k footage from my Sony. By comparison, the Mac Mini i7 6 Core is unusable on it - the lag is really, really irritating.
The 13" exported from FCPX faster than my iMac Pro too.
It was very nearly the last Apple product I ever bought!
[automerge]1589564792[/automerge]
Surprisingly good! I was happily using it for FCPX on 4k footage from my Sony. By comparison, the Mac Mini i7 6 Core is unusable on it - the lag is really, really irritating.
The 13" exported from FCPX faster than my iMac Pro too.
Yes I think if it's doing that, your not running it outside in the middle of a desert so outside of operating temp then there's a fault. The point of that base clock is that's the worse case you should ever expect when loading the machine up. No sensible tech company though probably puts in just enough cooling that you only just hit the base clock, they likely want a nice margin so they avoid a lot of returns.
The MacBook Air of course might be an interesting case there considering it's cooling and it running at 100C so easily and often. I've not watched enough of those videos to see but maybe that one doesn't go much above base clock under heavy load because of the poor cooling solution.
So much misinformation has been written and broadcast about the MacBook Air. It is not that it performs more slowly because it has a “poor cooling solution.” It has a very cleverly designed, but lower capacity cooling system because it uses a processor that is slower processor by design. Hook up the MacBook Air to the world’s best cooling system, and it won’t run significantly faster than it does. Intel designed the chip to run at 12W max in sustained tasks (10W base). By contrast, the base chip in the 13” Pro runs at 15W, with a max of 25W. That’s why the latter chip is almost twice as fast at CPU-intensive tasks.
I'm going to have to disagree with you a little on this. If the CPU is hitting 100C which is the thermal limit then it's throttling because of the temp. It's unlikely that it's hitting 100C at exactly the same time it's drawing it's maximum power draw, so more performance would have been possible from the same chip with better cooling.
There's plenty of benchmarks you'll find on the Internet of laptops with the same CPU's where one could be noticeably faster than another simply because the cooling solution was better. Case in point the 16in MacBook Pro will come out faster than the 15in it replaced despite them having the exact same CPU because the 16in has better cooling.
The 13in of course looks to be hitting 100C on the 10th gen so again with better cooling still it might have performed better also. Most thin laptops of course have the same problem, it's just a lot worse in the Air.
MaxTech actually did a test where he hooked up the Air CPU to a liquid cooler. It performed a little bit faster, but it still topped out around 1.7 GHz in sustained tests. Remember, it is running the Y-series processors, which were originally intended for fanless computers like the 12” MacBook. Apple added a basic fan to the MacBook Air, but even if you put that processor in a 16” MacBook Pro, it won’t perform that much faster than it does. Apple has designed the Air so that the chassis is cool even if the CPU is at 100C. They don’t use the case for cooling (as some of the hacks on that 45+ page thread on the Air forum have done). Stated otherwise, they aren’t going to go great lengths to extract maybe another 10% out of the CPU at the expense of a less comfortable notebook.
The 13” Pros don’t quite hit 100%. They hit the high 90s.
Yes I'm aware it has a Y series processor. I didn't know about the liquid cooler test but that's interesting. Seems sensible that Apple would have made sure they were at least close to cooling it as good as it needs but they are obviously off a little bit.
The 10th gen 13 can hit 100C. I've not seen a video of the 8th gen going quite as high but it also doesn't go quite as high on the wattage.
Check this video about 3:30 in to see it: Tally Ho MacBook Pro 13 First Look
Apple’s existing cooling system works fine for everyday usage, and I’m sure that’s why Apple designed it as they did. The 10% difference is on things like exporting videos that really tax the CPU.
I'm not saying it doesn't as most people will never know it's hitting 100C anyway. It could have been a bit better but probably not without more cost and it's the lower end device. So if it's good enough, that's likely fine vs a more expensive better solution. It's hitting a price point at the end of the day.
That liquid cooling video does highlight what removing heat from the 2020 MacBook Air can do for the i5 CPU.
It ran Cinebench R20 faster by 19% with liquid cooling, and the CPU temperature dropped from 100 degrees to 51 degrees C. So instead of hitting a thermal limit, it hit a power limit. Apple sets those power limits in firmware. So, even if it runs at 51 degrees C, the processor will not draw additional power without a change in Apple's firmware settings.
Last edited:
I think the point about 'good enough' is well made and often gets lost in discussions by us enthusiasts. I think that real world design and development processes are based on a very different mindset to the one we bring to these kinds of discussions and are rooted in a mindset of 'good enough' (for intended use) rather than 'best possible'.
I'd rephrase your comment above slightly and suggest that it's not necessarily a cost / price point issue, so much as one rooted in the different design goals of the MBA & MBP. Given the intended use cases for the MBA, I can easily imagine Apple deciding to leave <=19% of potential Ghz performance on the table. The 10w 10th gen 'Y' chips in the MBA deliver a 'good enough' snappy performance for 99% of its intended real world uses (single threaded and short, bursty multi threaded workloads), and the machine remains cool to the touch, thin etc. while doing so.
From reports, even the 10th gen MBP 13" is hitting the 100degC T junction on longer runs. Again, it's leaving potential Ghz on the table which is also no doubt another 'good enough' design decision by Apple.
[automerge]1589630640[/automerge]
Its also the case that the way Intel produce and market their chips inevitably encourages this kind of speculation. Rather than the 'this i5 chip runs at x speed and the i7 runs at x+n speed' of earlier times, Intel now market chips in which the stated speed is pretty nominal and the boost speed is widely variable depending on the specific power, temp and cooling parameters instigated by oems...
Last edited:
The MBA's cooling is barely enough, but still stinks for the quad core CPU. Notebookcheck just did a review on the i3 and the i3 is running at ambient noise most of the time, while the i5 under the same load quickly hits 45db in noise. While you might not be leaving much performance on the table from power limits, the quad core i5/i7 MBA's should at least be running close to the i3 in moderate loads, not over 13db louder. For some that noise level is barely perceptible, but was quite noticeable for me on the i7 2020 MBA. As I mentioned on another thread, the MBP does run quieter on the same loads (photo editing/installing software), but it definitely gives off much more heat.
My comment was really more about the sense of proportion - of 'good enough' - introduced in the posts above which is a key consideration in real world design processes that tends to get overlooked in discussions on forums like this. I wasn't trying to rehash the arguments about the MBAs thermals / performance / fan noise that's been done to death elsewhere.
In your particular example my takeaway would simply be that your use case (batch photo processing?) is better suited to a MBP. And to this end you recently swapped your MBA for a MBP, I think? We tend to universalise our experiences; your use case ended up being better met by a machine with different design goals (the MBP) but I don't think one has to conclude that the design of the MBA has failed for its intended use cases on the strength of this.
Notebookcheck released their review of the $1799 2020 MBP 13 with 4 TB ports.
Compared to the 8th-gen CPU in the 2019 MBP 13 with 4 TB ports, the 2020 model's 10th-gen CPU scored 4% higher in Cinebench R20's single-core test and 16% higher in the multi-core test.
For graphics, the 2020 model scored 33% higher in 3DMark (2560x1440 Time Spy Graphics) than the 2019 model.
Fan noise is slightly higher on average in the 2020 model when under load (34.4 dbA vs. 33.9 dbA).
The 2020 runs about 20 degrees hotter under sustained CPU and GPU load.
The 2019 can sustain 2.3 GHz in all four cores using 32W at 79 degrees C.
The 2020 can sustain 2.4 GHz in all four cores using 34W at 99 degrees C.
Compared to the 8th-gen CPU in the 2019 MBP 13 with 4 TB ports, the 2020 model's 10th-gen CPU scored 4% higher in Cinebench R20's single-core test and 16% higher in the multi-core test.
For graphics, the 2020 model scored 33% higher in 3DMark (2560x1440 Time Spy Graphics) than the 2019 model.
Fan noise is slightly higher on average in the 2020 model when under load (34.4 dbA vs. 33.9 dbA).
The 2020 runs about 20 degrees hotter under sustained CPU and GPU load.
The 2019 can sustain 2.3 GHz in all four cores using 32W at 79 degrees C.
The 2020 can sustain 2.4 GHz in all four cores using 34W at 99 degrees C.
This video, by itself, almost convinces me that I want the 13.3" version and do not need to consider the 16" version.
This is beyond disappointing. Intel messed up the 10nm big time. Its understandable why they didn't adopt 10nm for the bigger chips yet — they would have probably been slower...
Agree, lower speeds, same or worse power consumption, only GPU has significant gains. I guess Tiger Lake will be the presumptive true 10nm chip with supposedly significant gains in CPU/GPU...