15 inch 2017 buying advice needed. Please help

mario-64

macrumors regular
Original poster
Jun 23, 2012
218
103
Hi all,

I'm looking to purchase a 2017 15" MBP soon for use with CPU intensive applications. Before I plunk down such a huge amount of money I want to be sure I pick the right one. I'm eyeing the 3.1ghz model but have read that all the models throttle considerably under load. Given that, would the higher clocked CPU actually be faster in real world use? Would the slower CPU actually end up being faster?

Thanks,
Mario
 

monkeydax

macrumors 6502
Nov 2, 2012
355
91
Don't think the slower CPU will actually end up being faster, but I believe the improvement in speed you get from the highest end CPU is just incremental. I'd say just get the 2.9 GHz model as the other just isn't worth the increment.
 
  • Like
Reactions: Naimfan

ZapNZs

macrumors 68020
Jan 23, 2017
2,310
1,150
Could you tell us a bit more about your usage? For example, the actual applications used (and if they are obscure apps whether they are single/multithreaded), whether this will be more of a burst versus sustained load, if this is mainly CPU-intensive or CPU + GPU intensive, and if the price of the CPU upgrade is financially justifiable to you for a relatively small performance gain?
 
Last edited:

mario-64

macrumors regular
Original poster
Jun 23, 2012
218
103
Thanks to you both. From a gaming perspective it will be WoW, which is very CPU dependent and would be for sustained periods. Otherwise it will be video editing with Final Cut Pro
 
Last edited:

ZapNZs

macrumors 68020
Jan 23, 2017
2,310
1,150
What GPU/SSD specifications are you considering?

I ask because of the $100 difference between the 7700 and 7820, and an additional $200 difference between the 7820 and 7920 (or a $300 difference between the 7700 and 7920) is obviously enough money to upgrade other aspects of the MBP. For such usage, it sounds like the 560 GPU upgrade will have the biggest impact for you with day-to-day usage, and that the larger 1TB or 2TB SSD (slightly faster write speeds, more flexibility to keep ongoing projects on the local machine, better system performance & longer SSD life thanks to a less-filled drive) will have the second biggest impact, followed then by CPU. If that is the case, the 2.9 CPU/560 GPU/1TB SSD might be a solid match for your needs: you'd still get the larger L3 cache (which I vaguely recall reading is something that was found to have benefits with WoW), and this would give you $200 towards the SSD/GPU upgrade over opting for the priciest CPU option...from what I have read, it appears in benchmarks that each CPU upgrade adds about 5% overall performance to the tier below, and that applies to both single-core and multicore performance (someone smarter than me could verify this.)

Someone who knows more about CPUs than I do could verify this, but it is possible the 7820 and the 7920 are identical chips, and the 7920s are just higher binned models? If so, they could discuss what the implications are in terms of thermal efficiency. In the case of the retina MacBook, where the m3/i5/i7 are all the same chips, the gains made there between the base m3 and flagship i7 are very, very limited - relative to the pricing of the upgrades.
 

mario-64

macrumors regular
Original poster
Jun 23, 2012
218
103
Great thanks again. I can get the 2.9, 560, 1TB for $2879 vs the 3.1 with only 512gb for $2899. 3.1 1TB jumps to $3299. I really want as much CPU power as I can get but the reports of throttling on sustained loads has me hesitating on the 3.1. I’d hate to spend that money and end up with a system that’s actually slower
 

TofSanity

macrumors 6502a
Oct 29, 2010
660
85
Don’t mean to highjack this thread but the title says exactly what I’m in need of lol

I recently sold my 2016 15” MacBook Pro. I had the mid level 2.7, 512gb & with 450 graphics. My question is, besides the SSD size, is the entry level 2017 15” a bit better than the one I had? Thanks
 

leman

macrumors G4
Oct 14, 2008
10,002
4,565
I've tested a bunch of 15" models with different CPUs and none of them showed any kind of throttling under prolonged heavy CPU load. I also didn't observe any throttling when both CPU and GPU were taxed at the same time. Yes, you'd might get slightly lower turbo frequencies compared to under what you get in burst situations. Thats not throttling.
[doublepost=1509231723][/doublepost]
I recently sold my 2016 15” MacBook Pro. I had the mid level 2.7, 512gb & with 450 graphics. My question is, besides the SSD size, is the entry level 2017 15” a bit better than the one I had? Thanks
There is barely any difference. But the 2017 has a better GPU, since they made the mid-tier one standard.
 

TofSanity

macrumors 6502a
Oct 29, 2010
660
85
I've tested a bunch of 15" models with different CPUs and none of them showed any kind of throttling under prolonged heavy CPU load. I also didn't observe any throttling when both CPU and GPU were taxed at the same time. Yes, you'd might get slightly lower turbo frequencies compared to under what you get in burst situations. Thats not throttling.
[doublepost=1509231723][/doublepost]

There is barely any difference. But the 2017 has a better GPU, since they made the mid-tier one standard.
That’s what I was thinking. I don’t need anything more than I had before. So I was hoping to just purchase the entry MacBook Pro. Thanks for the info
 

mario-64

macrumors regular
Original poster
Jun 23, 2012
218
103
Thanks leman. I certainly hope that’s true but there are a few YouTube videos and even posts here at Macrumors which say that the CPUs regularly hit thermal limit (100c) and throttle down since Apple values quietness over performance. So your testing has shown this not to be true? If so I may go for the 3.1 then
 

Melbourne Park

macrumors 6502
Mar 5, 2012
281
31
Definitely getting the 560. My questions are concerning the CPU and potential throttling
I reckon contact Apple and ask to trial a high end machine, with your software on your disk. Or if you are in the USA, or many Apple stores world wide, buy the fast unit and if its throttling, then return it within two weeks. If the top model throttles maybe the slower one will not? Only way to find out is to either test, or buy and try.
 

Melbourne Park

macrumors 6502
Mar 5, 2012
281
31
A slight difference and probably not worth the money ....

Correct from Geekbench, Mult CPU test showed the 3.1 being 2% faster than the 2.9, while 4% faster in single thread performance.

MacBook Pro (15-inch Mid 2017)
Intel Core i7-7920HQ @ 3.1 GHz (4 cores)
15667


MacBook Pro (15-inch Mid 2017)
Intel Core i7-7820HQ @ 2.9 GHz (4 cores)
15313

Single CPU a bit of a difference though:


MacBook Pro (15-inch Mid 2017)
Intel Core i7-7920HQ @ 3.1 GHz (4 cores)
4649


MacBook Pro (13-inch Mid 2017)
Intel Core i7-7567U @ 3.5 GHz (2 cores)
4605


MacBook Pro (15-inch Mid 2017)
Intel Core i7-7820HQ @ 2.9 GHz (4 cores)
4493
 
  • Like
Reactions: Queen6

Naimfan

Suspended
Jan 15, 2003
4,669
1,996
A slight difference and probably not worth the money ....

Correct from Geekbench, Mult CPU test showed the 3.1 being 2% faster than the 2.9, while 4% faster in single thread performance.

MacBook Pro (15-inch Mid 2017)
Intel Core i7-7920HQ @ 3.1 GHz (4 cores)
15667


MacBook Pro (15-inch Mid 2017)
Intel Core i7-7820HQ @ 2.9 GHz (4 cores)
15313

Single CPU a bit of a difference though:


MacBook Pro (15-inch Mid 2017)
Intel Core i7-7920HQ @ 3.1 GHz (4 cores)
4649


MacBook Pro (13-inch Mid 2017)
Intel Core i7-7567U @ 3.5 GHz (2 cores)
4605


MacBook Pro (15-inch Mid 2017)
Intel Core i7-7820HQ @ 2.9 GHz (4 cores)
4493
Right. I defy anyone to notice the difference in real use.
 

mario-64

macrumors regular
Original poster
Jun 23, 2012
218
103
A slight difference and probably not worth the money ....

Correct from Geekbench, Mult CPU test showed the 3.1 being 2% faster than the 2.9, while 4% faster in single thread performance.

MacBook Pro (15-inch Mid 2017)
Intel Core i7-7920HQ @ 3.1 GHz (4 cores)
15667


MacBook Pro (15-inch Mid 2017)
Intel Core i7-7820HQ @ 2.9 GHz (4 cores)
15313

Single CPU a bit of a difference though:


MacBook Pro (15-inch Mid 2017)
Intel Core i7-7920HQ @ 3.1 GHz (4 cores)
4649


MacBook Pro (13-inch Mid 2017)
Intel Core i7-7567U @ 3.5 GHz (2 cores)
4605


MacBook Pro (15-inch Mid 2017)
Intel Core i7-7820HQ @ 2.9 GHz (4 cores)
4493
Thanks Melbourne. So basically a 4% difference I see.
 

Melbourne Park

macrumors 6502
Mar 5, 2012
281
31
Another interesting thing is the speed of older macbook pro computers - here are the multi-thread results:

2015 is 6% slower, but a 2105 2.8 ghz is no slower than a 2017 2.8 ghz. Hmmm ... seems the 2.9 Ghz 2017 is worth it, the others are not.

Of course, the only way is to run your application .... that's the only real test.

Makes me think though that (the reason I know this stuff is that I want a new macbook) - a refurb 2.8 Ghz Macbook Pro is not a bad deal for me as I get a nice keyboard, a photo card slot and I won't need to buy other dongle or adapters either.

Another thread says that next year 6 threaded CPUs may ship and then, there will be a major jump in performance ...

Incidentally since the site has two different speeds for the 2015 Mid 2,8 Ghz units, perhaps there's an error in the site?


MacBook Pro (15-inch Mid 2017)
Intel Core i7-7920HQ 3.1 GHz (4 cores)
15667


MacBook Pro (15-inch Mid 2017)
Intel Core i7-7820HQ @ 2.9 GHz (4 cores)
15313


MacBook Pro (15-inch Retina Mid 2015)
Intel Core i7-4980HQ @ 2.8 GHz (4 cores)
14435


MacBook Pro (15-inch Mid 2017)
Intel Core i7-7700HQ @ 2.8 GHz (4 cores)
14421


MacBook Pro (15-inch Retina Mid 2015)
Intel Core i7-4980HQ @ 2.8 GHz (4 cores)
14316
 

PBG4 Dude

macrumors 68030
Jul 6, 2007
2,646
1,810
The 2.9GHz CPU has the same 8MB L3 cache as the 3.1GHz CPU (vs 6MB on base CPU), so you’re getting the most important CPU improvement in the 2.9GHz CPU.