RAM and CPU options for 2017 Macbook 12 in

[Mike Boreham]

Hey, I have done this test when comparing the maxed out 12-inch MacBook to 13-inch MBP a while ago. I didn’t get any throttling issues at all. Take a look at screenshot or check out this page https://eldargezalov.com/news/12-inch-macbook-vs-13-inch-macbook-pro

Thanks Eldar...very interesting and nice review!
I am not at all surprised that the sudden dips seen on the Notebookcheck review have not occurred,

But I am surprised at the low level of fall off you show, ie almost none. When I have done the repeated Cinebench tests on my 2016 M5 I have seen typically about 10%.

 

[EugW]

Good review. But again I wonder if there were background processes in that 2017 i5 test. My m3 scored slightly higher than the i5 in Cinebench. At some point I should try the repeated Cinebench test.

Is there a script that can automate this?

 

[Coolio1234]

Hey, I have done this test when comparing the maxed out 12-inch MacBook to 13-inch MBP a while ago. I didn’t get any throttling issues at all. Take a look at screenshot or check out this page https://eldargezalov.com/news/12-inch-macbook-vs-13-inch-macbook-pro View attachment 721552

Wow! Thanks for the validation man, feel so much better about my purchase
[doublepost=1506519934][/doublepost]

Good review. But again I wonder if there were background processes in that 2017 i5 test. My m3 scored slightly higher than the i5 in Cinebench. At some point I should try the repeated Cinebench test.

Is there a script that can automate this?

I think he accidentally listed his MacBook as an i5 when later on in the review it's actually an i7 (1.4ghz)

 

[EugW]

Wow! Thanks for the validation man, feel so much better about my purchase
[doublepost=1506519934][/doublepost]
I think he accidentally listed his MacBook as an i5 when later on in the review it's actually an i7 (1.4ghz)

Well, either way, my m3 on a single test scored higher at 262 for Cinebench. However, I have no idea how it will behave with repeated tests. Dunno if it it will throttle significantly.

If it does have to be tested manually for the repeated benchmark, maybe I’ll run it 8-10 times and stop there. Too tedious to do it more.
[doublepost=1506524065][/doublepost]Ah yes, you are correct. The Geekbench test confirms it is a Core i7-7Y75 at 1.4 GHz.

This doesn't bode well for the Core i7 then:

I will do my repeated tests tonight or tomorrow to see what happens, but I wonder if the Core i7 might even be throttling right from the beginning in the first test, cuz this is what I get with my first test on my 2017 Core m3:

My m3 scored 262 the only time I tried it (this time), and his i7 never ever broke 260 despite his having done it a few dozen times.

 

[Eldar Gezalov]

I think he accidentally listed his MacBook as an i5 when later on in the review it's actually an i7 (1.4ghz)

Hey, yes it is i7. it was a typo, thanks for pointing out.

My m3 scored 262 the only time I tried it (this time), and his i7 never ever broke 260 despite his having done it a few dozen times.

I think I saw it in other reviews as well, quite a few times. For some reason the score for i7 in Cinebench R15 test is on par or even lower than m3. I will do the test again to confirm.

 

[Mike Boreham]

For some reason the score for i7 in Cinebench R15 test is on par or even lower than m3. I will do the test again to confirm.

Just a thought...is it possible that your i7 was already warmed up for the Cinebench series you showed?

This would explain both the lower than expected initial result (compared to M3), and the low fall off in the tests after.

 

[Eldar Gezalov]

Just a thought...is it possible that your i7 was already warmed up for the Cinebench series you showed?

Even if I do a test right after powering my MacBook on, I always get 255~260 max. The lowest score I got during the repeated test was 235 points, which is about 10% drop.

My Geekbench results are almost always over 4300/8100.

Can anybody else do the simple Cinebench R15 test of maxed out MacBook?

 

[EugW]

2017 MacBook Core m3-7Y32 1.2 GHz - 25 runs of Cinebench R15.038. Yes I spent over an hour doing this.

Ambient room temperature 23°C. Table top was a granite breakfast bar. Running on battery power. Screen set not to dim. WiFi and Bluetooth off, and Sophos anti-virus turned off.

Note that part way through the 11th run, I moved the computer over a bit because I was rearranging the stuff on the counter. Low and behold, the score went up. I noticed the part of the table where the computer had been was a little bit warm (but not hot), so it seems moving the computer over to a cooler spot helped performance. So, right after run 22, I intentionally did the same thing again, and again the score went up.

Below are the raw scores for future reference in case someone wants to put it into an Excel spreadsheet, but the most valid scores are only the first 10 for the reasons stated above:

264
261
259
258
258
258
257
255
255
253
254
257
257
257
256
255
255
254
252
252
251
250
254
258
258

You'll note that not even once did the score drop below 250. It may have dropped a bit below 250 had I not moved the computer, but given that 25 runs is about 65 minutes, that means 10 runs is around 25 minutes of sustained load with only about 1 second or so gaps between runs. I'd say I'm pretty happy with that result. In other words, it would have taken half an hour of sustained load to drop below a score of 250 on this mighty m3.

Compare this to Notebookcheck's 2017 results. I'm pleased to say I did not replicate their results:

Furthermore, my 2017 Core m3 handily beats all three MacBook CPU tiers from Notebookcheck's 2016 MacBook face-off:

And as mentioned, this surpasses Eldar's 2017 Core i7 for whatever reason:

 

[Mike Boreham]

Well done and well reported thanks!

You seem to have a good M3 there!

At least from memory the Geekbench scores for the M3 i5 and i7 are in the expected order.

Would be interesting to find out the spread of M3 results available.
[doublepost=1506575447][/doublepost]



mmmm....just turned wifi and bluetooth back on and got 244 again so cancel the above!

mmmm again.....just ran Cinebench again (wifi and BT off) and got 244 again. Ran it six more times and got exactly 244 each time. Then I quit Cinebench and relaunched it and got 215. Obviously it had got stuck which doesnt seem to have happened to anyone else so far, but beware if you get a string of identical results.

 

[Quddy]

I've been using the Early 2016 Macbook m3 (with Windows 10 Pro via Boot Camp) for over a year now and overall very pleased with it. The only problem I've had is sometimes it slows down quite significantly, usually when it's plugged into an external monitor and charging at the same time.

I guess this is due to throttling as I live in SE Asia and the ambient temp is usually 28-30°C indoors. I was thinking maybe to upgrade to a 2017 i5 Macbook... Would this be worth it or would I get the same results due to the temps here? Or maybe wait for the refresh later this year for an even bigger jump in performance?

 

[Mike Boreham]

I've been using the Early 2016 Macbook m3 (with Windows 10 Pro via Boot Camp) for over a year now and overall very pleased with it. The only problem I've had is sometimes it slows down quite significantly, usually when it's plugged into an external monitor and charging at the same time.

I guess this is due to throttling as I live in SE Asia and the ambient temp is usually 28-30°C indoors. I was thinking maybe to upgrade to a 2017 i5 Macbook... Would this be worth it or would I get the same results due to the temps here? Or maybe wait for the refresh later this year for an even bigger jump in performance?

In your shoes I would want to understand exactly what causes it to slow down. You said "sometimes"...what are you doing when it happens. You could try running Cinebench repeatedly like above to see if you can see the slow down reproducibly.

If you aren't sure what is causing the slow down you may find the 2017 M5 does the same, although being generally faster it should be better.

Out of interest which external monitor are you using it with?

 

[Quddy]

In your shoes I would want to understand exactly what causes it to slow down. You said "sometimes"...what are you doing when it happens. You could try running Cinebench repeatedly like above to see if you can see the slow down reproducibly.

If you aren't sure what is causing the slow down you may find the 2017 M5 does the same, although being generally faster it should be better.

Out of interest which external monitor are you using it with?

Usually just browsing or Youtube in Firefox. Chrome is a little bit better. Launching the app starts taking longer, opening new tabs and everything in general starts becoming laggy.

I have the j5create dongle and connect via HDMI usually to a monitor (HP Pavilion 25xi 1080p) and sometimes to flatscreens TVs also 1080p. Only single display not dual so it should in theory require less processing power compared to the Macbook display. Maybe there is some USB overhead? I also have a wireless mouse and keyboard connected to the dongle.

It's ok when I first plug the Macbook into the dongle but then gradually begins to slow down. Rebooting doesn't help but unplugging from the mains helps a lot so I assumed it was down to heat from charging in combination with using the dongle.

 

[Mike Boreham]

Usually just browsing or Youtube in Firefox. Chrome is a little bit better. Launching the app starts taking longer, opening new tabs and everything in general starts becoming laggy.

I have the j5create dongle and connect via HDMI usually to a monitor (HP Pavilion 25xi 1080p) and sometimes to flatscreens TVs also 1080p. Only single display not dual so it should in theory require less processing power compared to the Macbook display. Maybe there is some USB overhead? I also have a wireless mouse and keyboard connected to the dongle.

It's ok when I first plug the Macbook into the dongle but then gradually begins to slow down. Rebooting doesn't help but unplugging from the mains helps a lot so I assumed it was down to heat from charging in combination with using the dongle.

Difficult to comment when so much is different from my usage.

I suggest you install Intel Power Gadget and leave it running. It shows CPU and Temperature (amongst others). Get used to what it shows when operation is normal and see what is happening when it slows down.....it should be obvious from IPG.

You can download it here:

https://software.intel.com/en-us/articles/intel-power-gadget-20

Might also be interesting to spend some time in macOS doing the same things and see if there is a Bootcamp factor....wouldn't be surprising.

I have just looked back at some Cinebench tests I did when I had Bootcamp installed on my MacBook and it scored higher than in macOS. At that time my best macOS was 236, and Bootcamp Win10 was 262, all with my 2016 M5. The same was true on my 15" rMBP (579 vs 668). I didn't do any repeat running to assess fall off.

 

[EugW]

I reposted my Cinebench results as a new thread:

https://forums.macrumors.com/thread...-during-1-hour-of-sustained-cpu-load.2073415/

 

[Coolio1234]

Here you go! Very impressed with my results. I must admit, last night when I got the mac and started running the tests, I was only getting 179, 214, etc. I'm assuming that's because of the fact that spotlight indexing may have been taking place, or even me installing High Sierra beforehand had a role in it too. This morning, I just opened up my Mac (and even accidentally left finder, a safari tab, mail open) and ran the test at least 10 times.

 

[EugW]

Here you go! Very impressed with my results. I must admit, last night when I got the mac and started running the tests, I was only getting 179, 214, etc. I'm assuming that's because of the fact that spotlight indexing may have been taking place, or even me installing High Sierra beforehand had a role in it too. This morning, I just opened up my Mac (and even accidentally left finder, a safari tab, mail open) and ran the test at least 10 times.

They seem to converge, although it's hard to judge after run 10, since in run 11 the i7 got a phone call, and the m3 was moved to a different part of the table.

 

[Coolio1234]

Right. I'll see if I can get a more consistent test soon.

I kept my laptop in the same place for the tests, btw.