Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Thanks! Your scores are really stable. And consistently high. The potential battery/ac thing is weird...Maybe it depends on whether the AC is charging the battery (vs it being full)? I.e. If the battery is half full and charging, maybe that somehow limits power delivery available to the CPU?
That's possible, and I read that the higher tier 2020 13" has voltage ratings to use either the 61w or 87w adaptor - maybe that would have some impact, might be worth someone checking who has access to both.

https://www.macrumors.com/2020/05/06/macbook-pro-higher-power-adapters/
 
In case anyone is interested, Geekbench have added the new 2020 MBPro to their global Benchmark charts now:

The 10th-gen i7 is now the second fastest mac according to GB Single core scores:
A good 100 points (10%) faster than both the 10th-gen i5 and the 8-core i9 16 inch machine (despite its 5 Ghz turbo capability!).

I am still rather sceptical of these averages though, given all the variation observed above.

Interestingly, the 2020 8th gen is placed well down the pack in single core despite its very strong reviews it receives:
1591198314117.png
 
  • Like
Reactions: iRun26.2
@Spectrum More like 8.12% faster single core and 6.8% faster multicore, no? ;)

Single core performance for both is impressive, considering the i5 rivals a desktop i5-9600k. And the MBP i5 and i7 are about Apple's fastest quadcore chips, rivaling the last quadcore iMacs.

Value, though, is another question: Amazon and Best Buy recently had $200 off the 10th gen i5s but not the i7s. I can almost justify the i7 for $200 more, but when the i5 is $400 cheaper it's a no brainer. (I grabbed a used-like new one for another $90 off that price.) That might change if discounts come to the i7 too, but I decided not to wait for that . . .
 
  • Like
Reactions: KPOM and Spectrum
@Spectrum More like 8.12% faster single core and 6.8% faster multicore, no? ;)

Single core performance for both is impressive, considering the i5 rivals a desktop i5-9600k. And the MBP i5 and i7 are about Apple's fastest quadcore chips, rivaling the last quadcore iMacs.

Value, though, is another question: Amazon and Best Buy recently had $200 off the 10th gen i5s but not the i7s. I can almost justify the i7 for $200 more, but when the i5 is $400 cheaper it's a no brainer. (I grabbed a used-like new one for another $90 off that price.) That might change if discounts come to the i7 too, but I decided not to wait for that . . .
Yeah. I was rounding rather generously!

Another way to look at this is that year-on-year single core CPU performance increases ~10% or so on equivalent machines (Exceptions are changes from dual to quad/hex/Oct etc).

So, for ~10% extra price now, the i7 will actually last for an additional year in terms of relative speed.
This means, to any arbitrary end user, an i7 will start to feel "slow" (compared to a new machine) about a year later than will the i5.

+10% price ($200) is good value if it means the computer is good enough for 5 vs 4 years, AND is faster (if only fractionally) for the entire duration of that time.
 
Last edited:
  • Like
Reactions: iRun26.2
Got this with around 1.5-2% base CPU usage.

1591385259402.png

Honestly it's fine, I don't really care if it isn't 4500. I haven't noticed any limitations on speed or wished it was faster. Higher single-core score than the Mac Pro though, damn. That's what's more important to me I think, anyway.
 
  • Like
Reactions: Spectrum
Got this with around 1.5-2% base CPU usage.

View attachment 921844
Honestly it's fine, I don't really care if it isn't 4500. I haven't noticed any limitations on speed or wished it was faster. Higher single-core score than the Mac Pro though, damn. That's what's more important to me I think, anyway.
Is that an i5 or i7?

Out of interest...can you retry on battery power? (Unless this already was.)
 
Last edited:
For SC scores, it is interesting that both i5 and i7 show a bimodal distribution...with peaks about 150 points apart.

I wonder if that could be the AC power versus battery difference? Or maybe 16 vs 32 GB RAM?

I caught a comparison review on 'tube of the i7 MBP 13" vs the Dell XPS 13 9300 (2020 model) also with an i7 and turns out the MBP 13" runs the same on AC power as on battery under heavy load and won out across the board.


This thing is such a beast! I7/32/1TB just received yesterday.

View attachment 917946

I can never really understand why GB4 I got similar spec's on the 2018 MBP 13" i5 model I had (TB with 4 TB3 ports) as people are now getting with GB5. Due to this I don't fully trust the 'benchmark' scores since the benchmark keeps changing every year, using a very low spec machine. Unfortunately lower spec machines are now with slower RAM, lower wattage cpu's and even slower PCIe storage.

At least Apple compares with the outgoing model so you have a good real-world assessment to count on. Yes I know so many trust GB.
 
GB is fine for what it is, but is less good for estimating sustained multi core CPU activity.

However, you cannot compare GB4 with GB5. They are an updated set of tests, and are calibrated against a different baseline. This is done on purpose to *prevent* people comparing scores from GB3 vs GB4 vs GB5 etc.

For me, I feel it is roughly time to upgrade if either the feature set of a computer offers something I want (new keyboard for example, better Ports, more RAM or SSD capacities), or the CPU/GPU offers 1.5-2x improvement in something like Geekbench, because this is a good indication that I will notice a real-world improvement.

For example I went from 2011 quad i7 mini to 2018 i7 mini (2-3x score increases). And I’m looking to upgrade from a 2014 13 inch MBPro i7 to a 2020 i5 or i7 (~2x increases.)

However, the latter decision is more of a tricky decision for me because I will miss the ports and Magsafe that the 2014 comes with.

I caught a comparison review on 'tube of the i7 MBP 13" vs the Dell XPS 13 9300 (2020 model) also with an i7 and turns out the MBP 13" runs the same on AC power as on battery under heavy load and won out across the board.




I can never really understand why GB4 I got similar spec's on the 2018 MBP 13" i5 model I had (TB with 4 TB3 ports) as people are now getting with GB5. Due to this I don't fully trust the 'benchmark' scores since the benchmark keeps changing every year, using a very low spec machine. Unfortunately lower spec machines are now with slower RAM, lower wattage cpu's and even slower PCIe storage.

At least Apple compares with the outgoing model so you have a good real-world assessment to count on. Yes I know so many trust GB.
 
@Spectrum Wondering if you'd know something. The i5 is rated to turbo boost up to 3.8GHz, but I've only ever seen it go up to a sustained 3.3GHz. It's never gone beyond that, as confirmed by iStat Menus.

Is my CPU underclocked or something? I'm pretty sure I should be getting up to 3.8 at least in bursts, as per the tech specs.
 
@Spectrum Wondering if you'd know something. The i5 is rated to turbo boost up to 3.8GHz, but I've only ever seen it go up to a sustained 3.3GHz. It's never gone beyond that, as confirmed by iStat Menus.

Is my CPU underclocked or something? I'm pretty sure I should be getting up to 3.8 at least in bursts, as per the tech specs.

My understanding is that "Rated" turbo boost is single core mode. Multicore max is lower, and not usually quoted.
 
  • Like
Reactions: Patcell
@Spectrum Wondering if you'd know something. The i5 is rated to turbo boost up to 3.8GHz, but I've only ever seen it go up to a sustained 3.3GHz. It's never gone beyond that, as confirmed by iStat Menus.
Is my CPU underclocked or something? I'm pretty sure I should be getting up to 3.8 at least in bursts, as per the tech specs.
As @Mike Boreham said, it depends on the number of cores active. It is actually quite tricky to have only a single core active, since the CPU is often doing something else (even intermittently) in the background. One way to test is using Intel Power Gadget to monitor, and running a "Yes" test: Launch Terminal. Type: yes > /dev/null &
Spawn as many instances as you want to test cores and monitor max CPU frequency. Then kill processes with: killall yes

You may have to also adjust the sampling rate of Intel Power Gadget to observe the max CPU:
1591617023545.png


You can also use the inbuilt "testing" inside IPG:
1591617121678.png


In reality, the max CPU frequency you get will also be restricted by thermals. I don't know for sure, but based on the 1035G7 specs (copied below), I suspect the max 4-core turbo of the 1038NG7 is ~3.4 Ghz.
Similarly, based on the 1065G7 specs, the 4-core max turbo of the 1068NG7 is probably ~3.7 Ghz.

In practice I suspect these 4-core frequencies are possible for only very short periods due to thermals.


1591616469391.png


Also: If you are running any AVX code, the max CPU frequency drops down further due to them being very complex to execute (but very efficient).
 

Attachments

  • 1591617008392.png
    1591617008392.png
    83.3 KB · Views: 101
Last edited:
As @Mike Boreham said, it depends on the number of cores active. It is actually quite tricky to have only a single core active, since the CPU is often doing something else (even intermittently) in the background. One way to test is using Intel Power Gadget to monitor, and running a "Yes" test: Launch Terminal. Type: yes > /dev/null &
Spawn as many instances as you want to test cores and monitor max CPU frequency. Then kill processes with: killall yes

You may have to also adjust the sampling rate of Intel Power Gadget to observe the max CPU:
View attachment 922509

You can also use the inbuilt "testing" inside IPG:
View attachment 922510

In reality, the max CPU frequency you get will also be restricted by thermals. I don't know for sure, but based on the 1035G7 specs (copied below), I suspect the max 4-core turbo of the 1038NG7 is ~3.4 Ghz.
Similarly, based on the 1065G7 specs, the 4-core max turbo of the 1068NG7 is probably ~3.7 Ghz.

In practice I suspect these 4-core frequencies are possible for only very short periods due to thermals.


View attachment 922506

Also: If you are running any AVX code, the max CPU frequency drops down further due to them being very complex to execute (but very efficient).
Thanks for the info. Notebookcheck.net usually gives the turboboost possible on each number of cores, but the page on the 1038NG7 only lists maximum single core clock (3.8Ghz).

Do you by chance have a link to the macOS version of Intel PowerGadget? When I go to Microsoft's website there is huge list of possible downloads and I'm not sure which one is correct... thanks again.
 
Thanks for the info. Notebookcheck.net usually gives the turboboost possible on each number of cores, but the page on the 1038NG7 only lists maximum single core clock (3.8Ghz).

Do you by chance have a link to the macOS version of Intel PowerGadget? When I go to Microsoft's website there is huge list of possible downloads and I'm not sure which one is correct... thanks again.
Just about any of those files will work. I downloaded the first one on the list.
 
  • Like
Reactions: Patcell
Thanks for the info. Notebookcheck.net usually gives the turboboost possible on each number of cores, but the page on the 1038NG7 only lists maximum single core clock (3.8Ghz).

Do you by chance have a link to the macOS version of Intel PowerGadget? When I go to Microsoft's website there is huge list of possible downloads and I'm not sure which one is correct... thanks again.
The new 9th gen and 10th gen are a bit of a mess to know what to expect in terms of max turbo because the actual possible turbo is much more dependent on thermal constraints than it was for earlier CPU generations. People don't call this "throttling" because it is always above base speed, but it clearly is: i.e. The CPU is unable to reach its theoretical peak due to a lack of cooling. In older generations 4th, 5th, 6th gen, this was rarely true, and the CPU could almost always attain its maximum design spec turbo.


But remember that IPG itself will use up cores/resources...
 
Last edited:
Walk-in store (Adelaide - Australia) and got mine today: i7, 32gb ram, 2tb.
They had just received some 10th gen and the website still now does not even show them as available.
Not the first time I come to the conclusion it's faster to not order online and just go to the store...
[automerge]1589285407[/automerge]
I'm impressed with the ssd

I will be really happy when Apple Stores stock 32 GB systems in the store.
 
The new 9th gen and 10th gen are a bit of a mess to know what to expect in terms of max turbo because the actual possible turbo is much more dependent on thermal constraints than it was for earlier CPU generations. People don't call this "throttling" because it is always above base speed, but it clearly is: i.e. The CPU is unable to reach its theoretical peak due to a lack of cooling. In older generations 4th, 5th, 6th gen, this was rarely true, and the CPU could almost always attain its maximum design spec turbo.


But remember that IPG itself will use up cores/resources...
Thanks for the info and link!
 
I will be really happy when Apple Stores stock 32 GB systems in the store.
They always have. There's a "secret menu" of high end configs Apple stores carry. For example, the 2019 15 inch had a 2.4/32/560x/1tb or 2.4/32/v20/1tb and the 16 inch has 2.4/32/5500 8gb/2TB and now it seems they have a 13 2.3/32/2TB.

Edit: If you want to see for yourself, price out a Mac on Apple.com and put in a store location that has reopened, for example, Florida. Change the options until you see available today for pickup.

Screenshot 2020-06-08 10.52.47.png
 
Last edited:
Let me just chime in to post I've got quite consistent GB scores in my first week with the i7. Above 4800 after booting, a little lower, like 4600-4800, after having done just anything for some minutes or hours. Bought the i7 mainly for having a little more headroom with Logic Pro synths and plugins.

Bildschirmfoto 2020-06-07 um 21.36.37.png
 
  • Like
Reactions: ikkysleepy
I just got my i7 / 32 GB / 1 TB and I initially got 1361 and I updated the version of Geekbenchmark and now get 1196 after a reboot. I did enable file vault and added more apps since then but not sure if it was the apps or the updated Geekbenchmark version. Also, if anyone is waiting for their MBP and wants yet another unboxing video, check mine out.
 

Attachments

  • benchmark.png
    benchmark.png
    29.5 KB · Views: 102
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.