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

Max Yuryev tests the MBP 15" CPUs: 2.2 vs 2.6 vs 2.9

S

sennomulo

macrumors member
Original poster
Jul 18, 2018
81
102
Finally, someone got all three CPUs, the two stock model i7s plus the i9, and tested them against each other. This is the comparison I've been waiting for. The results are very interesting.

It's not a completely apples-to-apples test, since his 2.2 GHz model has the 555x GPU while the other two have the 560x GPU, but I guess it makes it a more relevant test for people looking at the two stock models.

 
  • Like
Reactions: Salaryman Ryan, Petetastic, uecker87 and 3 others
This is ridiculous. I’ve been with Apple since 2001 and remember how refreshing it was to move to OS X from windows. I need a new computer as my 2013 MacBook Pro is failing and for the first time in 17 years I’m actually considering a windows machine. Between the high price, high heat, and crappy (compared to the Nvidia cards in the pc’s) video cards it’s getting harder to justify the purchase.
 
  • Like
Reactions: duderman67, outsidethebox, pmore and 4 others
Yes..its surprising. The mid model i7 2.6 / 512 / 560X seems to be a better choice. Wish Best Buy in the US had carried the 32GB option could have used my gift cards+coupons to get it from them.
 
zshane1125 said:
Looks like the the best value / performance machine is the mid range i7 2.6 / 512 / 560X model.
Click to expand...
I've got this exact one. Factoring in the SSD and graphics card upgrade the price difference is not very bad (i.e. ~$84 for the 2.6GHz CPU in my country). CPU upgrades in the same class of device of the same generation is always pointless for a laptop.

I think the thermal situation and the general performance of Intel CPU won't see significant improvement until they can ship CPUs with 10nm architecture, now targeted to be in late 2019.

This year will be very interesting that the upcoming iPhone and iPad may actually exceed the Intel CPUs in Geekbench scores.
 
  • Like
Reactions: GuruZac, Patcell and M.Rizk
haruhiko said:
I think the thermal situation and the general performance of Intel CPU won't see significant improvement until they can ship CPUs with 10nm architecture, now targeted to be in late 2019.
Click to expand...

And it takes them a year or so to actually fill out all the mobile and desktop pieces at all the power variants. We will be well into 2020 before there’s chips for all of even Apple’s range (and that’s assuming Intel don’t continue their already 2+ year trail of broken promises)
 
  • Like
Reactions: haruhiko
All these tests people are doing are designed for desktop workstations and DO NOT work in a thin and light desktop. The turbo boost across all cores will be almost the same across the three processors.

These are not real world tests. For real world tests you need workloads that boost 2-4 cores as high as possible while relying on the remaining cores for intermittent assistance.

But such as an example would not make very good clickbait for youtubers who do not have long term sustainable career choices.
 
  • Like
Reactions: Macshroomer and bwintx
Runs BruceX, a GPU-dominated benchmark and wonders that the CPU performance doesn't matter :rolleyes:

People, the point that almost everyone is missing is that the definitive advantage of the i9 CPU is asymmetric BURST performance. Over high symmetric multicore sustained loads, all these CPUs will perform very similarly, since they will max at around 45-50W (as much as the MBP can dissipate). If you do video work, where loads are relatively long-lasting, memory access patterns are simple and algorithms are trivially parallel, go 2.6 model as best bang for buck. The i9 is a situational upgrade. I got it and I am keeping it, since the speedup for my work (programming, scientific algorithms prototyping etc.) is significant.

And yes, you are paying $300 more for 2-5% performance increase. Which has always been the case with the high-end CPU upgrades. You need to understand what you need, and buy accordingly. Price and performance are not in linear relationship!
 
Last edited:
  • Like
Reactions: Salty Pirate, Smeaton1724, mwb and 6 others
CreeptoLoser said:
All these tests people are doing are designed for desktop workstations and DO NOT work in a thin and light desktop. The turbo boost across all cores will be almost the same across the three processors.

These are not real world tests. For real world tests you need workloads that boost 2-4 cores as high as possible while relying on the remaining cores for intermittent assistance.

But such as an example would not make very good clickbait for youtubers who do not have long term sustainable career choices.
Click to expand...
So I'm really curious in what tasks my i9 would then show it's benefits. Myself, I often have to compile semi-large applications in X-code and Visual Studio, but I've seen the creator of Geekbench already stating that it didn't provide a benefit there. So if compiling and exporting video's are not it, what kind of other tasks would? I think it's far fetched to say 'Well they don't provide tangible performance benefits between them in any of the benchmarks we use, including those 'real-life' use-case tests that render relatively short video's, BUT they will in real life scenario's for which these laptops are used by real professionals.' I have seen you talk about 'laptop tasks' in other topics, but what are those?

Btw I'm not saying these CPU's are slow for laptops, they're really not, in fact my i9 scored 1124 in Cinebench and I found it to be fast enough for my expectations. However, when we're talking about which of the variations to pick out of these three, the performance difference between them just doesn't seem to match the price difference. You could just as well go for the 2.2 (or atleast the 2.6ghz) and notice close to no difference, saving $300-400.

I think at this point no one should buy the i9, period. If you want to do it for the burst just think about this: even when I limit everything to a single core through Instruments, the maximum speed it goes to is just under 4.4Ghz, which is less than 100Mhz difference compared to the 2.6Ghz Max Turbo boost. These Macbook cases are just not made for more. That doesn't mean they're not great machines, or that you shouldn't buy one; just that you shouldn't buy the i9 and expect better performance than the 2.6.

Honestly, I really can't see how anyone can advise the i9 at this point, without it being some form of trying of justifying their own purchase. I have the i9 but I won't do that.


Edit:
leman said:
The i9 is a situational upgrade. I got it and I am keeping it, since the speedup for my work (programming, scientific algorithms prototyping etc.) is significant!
Click to expand...
Do you have any tests that show the difference between the 2.6 and the 2.9 in these tasks? My 2.9 is certainly faster than last years model (and a lot quicker than my own 2014 model) I have no doubt about that, but I haven't seen any tests/results comparing the 2.6 against the 2.9 in these kind of tasks. Like I said above, I've only seen Geekbench's creator saying there was no difference.


Edit 2:
I will say that I don't really like the video as his 2.2Ghz didn't have the 560X. This leaves open a discussion point that shouldn't be there in these types of video's. If I was a Youtube creator and would want to create an objective video which shows the real difference, I would've had 3 completely similar models where the only difference was the CPU. I'm hoping there is one like that coming.
 
Last edited:
  • Like
Reactions: kael92 and haruhiko
Glad I returned my i9.
I went for a MBP 13" instead as my portable unit and am holding our for an iMAC instead now, which will be my main machine.....
 
  • Like
Reactions: ayyy_lmao
fokmik said:
If it will come in 2020 Apple will make for that 10nm a redesign to be even thinner
Click to expand...

I don't know, I think for 2020 Apple will probably focus on the thermal design, T2 chip, keyboard, FaceID and screen (4k) rather than making it thinner. I have made the decision to just wait out to see what the 10nm CPU laptops bring to the table be it macOS or Windows (so my XPS 9360 will need to hold out a couple of more years).
 
leman said:
Runs BruceX, a GPU-dominated benchmark and wonders that the CPU performance doesn't matter :rolleyes:

People, the point that almost everyone is missing is that the definitive advantage of the i9 CPU is asymmetric BURST performance. Over high symmetric multicore sustained loads, all these CPUs will perform very similarly, since they will max at around 45-50W (as much as the MBP can dissipate). If you do video work, where loads are relatively long-lasting, memory access patterns are simple and algorithms are trivially parallel, go 2.6 model as best bang for buck. The i9 is a situational upgrade. I got it and I am keeping it, since the speedup for my work (programming, scientific algorithms prototyping etc.) is significant.

And yes, you are paying $300 more for 2-5% performance increase. Which has always been the case with the high-end CPU upgrades. You need to understand what you need, and buy accordingly. Price and performance are not in linear relationship!
Click to expand...

Very good points here!

Do you think the i9 makes sense for photo editing? I work with very large files and sometimes I have to batch process hundreds of images...I think in my scenario I would actually benefit from the i9 BURST performance and money is not a problem at this point. (still using a 2012 MBP maxed out - 16GB RAM and 2 SSDs ... have been saving for a new laptop for a while)
 
shavou said:
Very good points here!

Do you think the i9 makes sense for photo editing? I work with very large files and sometimes I have to batch process hundreds of images...I think in my scenario I would actually benefit from the i9 BURST performance and money is not a problem at this point. (still using a 2012 MBP maxed out - 16GB RAM and 2 SSDs ... have been saving for a new laptop for a while)
Click to expand...

How long does the burst performance last? A couple of weeks ago people mentioned it throttles down after a couple of seconds or so, in which case you probably won't notice it in most scenario's if not all realistically speaking. I think the difference in performance might be <1%.
 
  • Like
Reactions: Aea and shavou
fokmik said:
Do you think the i9 makes sense for photo editing? I work with very large files and sometimes I have to batch process hundreds of images...I think in my scenario I would actually benefit from the i9 BURST performance and money is not a problem at this point. (still using a 2012 MBP maxed out - 16GB RAM and 2 SSDs ... have been saving for a new laptop for a while)
Click to expand...

I can't answer this, since I don't do photo editing. I think that photo editing is fundamentally similar to video editing though, just lighter. I'd say go for the 2.6 model.
 
  • Like
Reactions: shavou
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