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

cynics

macrumors G4
Jan 8, 2012
11,959
2,156
They matter if the final image still needs to go though the compositor, which it kind of does. High Sierra comes with a optimisation they call "Direct to Display" which avoids this, but its not on yet on 10.12

Unlike phones the monitor hardware (off GPU processing) processes the data sent to it. Easiest demonstration would be using a PC (or external monitor) if you are playing a game at 1080p on a 1080p monitor than unplug that monitor and plug in a 4K monitor but keep the same resolution the FPS won't change. Or even if the monitor is completely unplugged for that matter.
 

koyoot

macrumors 603
Jun 5, 2012
5,939
1,853
Wow so much mis-information in this thread.

An i5 desktop cpu isn't going to bottleneck a 560 ffs. What are we benchmarking at 1024x768?

We need people to actually benchmark and monitor clocks to see what is going on. Is the 560 throttling? Is it a driver problem/same drivers?

I have a 460 and if anyone want to compare with a 560 I can help them run through some benchmarks to see what is going on
It will not bottleneck RX 560 in GPU bound situation. But it will in CPU bound situation. Testing old games, in 1440x900 medium settings is CPU bound situation. The same goes for even 1080p medium settings.

For example, in medium Settings Radeon RX 560 in Overwatch in 1080p is able to push up to 120 FPS. Give it a bad CPU and it will bottleneck it. In 1080p Ultra settings in the same game there will be no difference if the RX 560 is paired with Pentium G4560, Core i7 4 core/8Thread, or Ryzen 7 8core/16 thread CPU. Because it is GPU bound scenario.
 
  • Like
Reactions: iKrivetko

ninja2000

macrumors 6502
Dec 16, 2010
338
75
It will not bottleneck RX 560 in GPU bound situation. But it will in CPU bound situation. Testing old games, in 1440x900 medium settings is CPU bound situation. The same goes for even 1080p medium settings.

For example, in medium Settings Radeon RX 560 in Overwatch in 1080p is able to push up to 120 FPS. Give it a bad CPU and it will bottleneck it. In 1080p Ultra settings in the same game there will be no difference if the RX 560 is paired with Pentium G4560, Core i7 4 core/8Thread, or Ryzen 7 8core/16 thread CPU. Because it is GPU bound scenario.
What you are saying makes sense. But we are talking an i5 3.8ghz desktop quad against a 3.8ghz mobile quad with hyper threading. The hyperthreading isn't going to make that much difference, it is more clock speed which is comparable. And in all honesty I expect the mbp quad to downclock under load unlike the iMac
 

jeanlain

macrumors 68020
Mar 14, 2009
2,465
959
They matter if the final image still needs to go though the compositor, which it kind of does. High Sierra comes with a optimisation they call "Direct to Display" which avoids this, but its not on yet on 10.12
Does it really make a difference? The compositing may induce some delay, it but should not reduce the fps by much.
BTW Leman, get you ass on the gaming forum and tell us how Metal 2 is great, pronto! ;)
 

leman

macrumors Core
Oct 14, 2008
19,530
19,709
BTW Leman, get you ass on the gaming forum and tell us how Metal 2 is great, pronto! ;)

There is a gaming forum? :D

I'm not sure that I'm competent enough to write a summary post on Metal 2, but if you have some specific questions, I'd be happy to answer them.

And yes, it's absolutely great. If I read this correctly it can now allow one to do some fairly crazy stuff beyond what DX12 and Vulkan are offering.
 

koyoot

macrumors 603
Jun 5, 2012
5,939
1,853
What you are saying makes sense. But we are talking an i5 3.8ghz desktop quad against a 3.8ghz mobile quad with hyper threading. The hyperthreading isn't going to make that much difference, it is more clock speed which is comparable. And in all honesty I expect the mbp quad to downclock under load unlike the iMac
Fully loaded MBP 2016 has 2.9 GHz Core i7 Skylake CPU. It Turbo's up to 3.6 GHz on all cores. Which means all 8 threads have 3.6 GHz. Compare this to All core Turbo boost on 4 threads of Core i5, which is the same.

35W TDP Core i7 7700T is able to maintain all core boost in 90% load scenarios.
 

jeanlain

macrumors 68020
Mar 14, 2009
2,465
959
There is a gaming forum? :D

I'm not sure that I'm competent enough to write a summary post on Metal 2, but if you have some specific questions, I'd be happy to answer them.

And yes, it's absolutely great. If I read this correctly it can now allow one to do some fairly crazy stuff beyond what DX12 and Vulkan are offering.
We have a couple of threads discussing WWDC announcement in the Mac & PC games forum. I'm personally wondering if argument buffers have equivalents in DX12 and Vulkan (if they correspond to reusable command buffers that people were asking), and if important features that some said were missing (sparse data, transform feedback, shader atomics on texture objects...) are now in Metal 2.
I could find answers myself but the Metal documentation is Chinese to me. :confused:
 

leman

macrumors Core
Oct 14, 2008
19,530
19,709
Unlike phones the monitor hardware (off GPU processing) processes the data sent to it. Easiest demonstration would be using a PC (or external monitor) if you are playing a game at 1080p on a 1080p monitor than unplug that monitor and plug in a 4K monitor but keep the same resolution the FPS won't change. Or even if the monitor is completely unplugged for that matter.

Well, did they benchmark the stuff on the external monitor? And yes, desktop composition is still a thing and it's on permanently on macOS.

Anyway, my point is that there might be a plethora of factors behind these performance differences. They certainly don't suggest that the 560 is slower than the 460
 

ninja2000

macrumors 6502
Dec 16, 2010
338
75
Fully loaded MBP 2016 has 2.9 GHz Core i7 Skylake CPU. It Turbo's up to 3.6 GHz on all cores. Which means all 8 threads have 3.6 GHz. Compare this to All core Turbo boost on 4 threads of Core i5, which is the same.

35W TDP Core i7 7700T is able to maintain all core boost in 90% load scenarios.

It can't maintain full boost on a mbp with sustained cpu and gpu load (I.e. Gaming) I have tested.

And hyperthreading won't give the cpu much of an advantage in games so you can see why this cpu is no superior to the desktop i5 in gaming and therefore the desktop i5 is not a bottleneck.
 

koyoot

macrumors 603
Jun 5, 2012
5,939
1,853
It can't maintain full boost on a mbp with sustained cpu and gpu load (I.e. Gaming) I have tested.

And hyperthreading won't give the cpu much of an advantage in games so you can see why this cpu is no superior to the desktop i5 in gaming and therefore the desktop i5 is not a bottleneck.
It is superior. Threads are always threads. Core i5 is loaded fully, while Core i7 is not fully loaded because of increased amount of work it can do. 8 Threads really that make that huge difference vs 4 threads. Obviously they are not physical cores, but the amount of work that is scheduled is higher on Core i7, thats why for high framerate scenarios everybody will tell you to buy Core i7 over Core i5.

If it does not sustain 3.6 GHz all of the time on all cores, to what level it declocks under load, then? 3.4? 3.5 GHz? It still will be faster overall for high framerate/low res/low detail scenarios.
 
  • Like
Reactions: Glideslope

jerwin

Suspended
Jun 13, 2015
2,895
4,652
It will not bottleneck RX 560 in GPU bound situation. But it will in CPU bound situation. Testing old games, in 1440x900 medium settings is CPU bound situation. The same goes for even 1080p medium settings.

For example, in medium Settings Radeon RX 560 in Overwatch in 1080p is able to push up to 120 FPS. Give it a bad CPU and it will bottleneck it. In 1080p Ultra settings in the same game there will be no difference if the RX 560 is paired with Pentium G4560, Core i7 4 core/8Thread, or Ryzen 7 8core/16 thread CPU. Because it is GPU bound scenario.
I suppose this page should be helpful, then..

http://www.techspot.com/review/1180-overwatch-benchmarks/page5.html
 

jeanlain

macrumors 68020
Mar 14, 2009
2,465
959
Geekbench is probably the one of the least useful tests to estimate GPU power. Right alongside cinebench and (gasp!) Xbench.
 
Last edited:
  • Like
Reactions: jerwin

jerwin

Suspended
Jun 13, 2015
2,895
4,652
The 580 has a geekbench compute of around 116K

http://browser.geekbench.com/v4/compute/818659

So only 8K better then a 575
I found a 580 that did 123856.
http://browser.geekbench.com/v4/compute/compare/818659?baseline=815105

maybe the extra RAM and i7 helped.
maybe the conditions of the test were different in other ways
[doublepost=1497203437][/doublepost]
Geekbench is probable the one of the least useful test to estimate GPU power. Right alongside cinebench and (gasp!) Xbench.
I suppose I could load up all the benchmarks on a flash drive and run them at the local Apple store 'til I get kicked out.
 

koyoot

macrumors 603
Jun 5, 2012
5,939
1,853
They are comparing effect of CPU bottleneck on fast GPU like GTX 1080.

You have proven my post, rather than disprove it.

TechSpot and Hardware Unboxed is run by the same guys.

And this is what they have to say about a difference in performance in different tiers of resolution and detail vs performance of the CPU.

I suggest to everyone watching this. GTX 1050 Ti is faster GPU than RX 560.

In High detail, high resolution the situation between Core i5 and Core i7 will be the same as is with GTX 1050 Ti. In low resolution/medium settings it will be the same situation as is with GTX 1080.
 
Last edited:

Kong1971

macrumors newbie
Jun 13, 2017
8
1
Metropolis, IL
Someone on Youtube finally benchmarked the 21.5 iMac playing games via bootcamp. The results were satisfactory to me. 1080p and pretty stable 60 FPS on most of the games they tried. I'm just a casual gamer and don't look to get 120 FPS or play at 4K. Just want to play Street Fighter V and Resident Evil here and there.

 

fokmik

Suspended
Oct 28, 2016
4,909
4,688
USA
Someone on Youtube finally benchmarked the 21.5 iMac playing games via bootcamp. The results were satisfactory to me. 1080p and pretty stable 60 FPS on most of the games they tried. I'm just a casual gamer and don't look to get 120 FPS or play at 4K. Just want to play Street Fighter V and Resident Evil here and there.

also the 27" with 570 or 575 i guess
 

zerozoneice

macrumors 6502
Jun 26, 2013
391
123
Someone on Youtube finally benchmarked the 21.5 iMac playing games via bootcamp. The results were satisfactory to me. 1080p and pretty stable 60 FPS on most of the games they tried. I'm just a casual gamer and don't look to get 120 FPS or play at 4K. Just want to play Street Fighter V and Resident Evil here and there.

buy a Nintendo?
 

TPNxl

macrumors newbie
Jun 7, 2017
3
1
Likely the REAL reason the Radeon Pro 560 sucks in the MBP 15" of mid-2017 is because of thermal throttling.

The equivalent 2016 MacBook Pro uses a Radeon Pro 460 (which I have). It has 16 CUs (1024 shaders) running at 850 MHz stock* with an 80 GB/s memory bandwidth to the 4GB of GDDR5. By the way, the TDP is 35 watts. It thermal throttles in most games that mostly max both CPU and GPU, especially in Windows.

The Radeon Pro 560 in the mid-2017 MBP, by comparison, has the same 16 CUs (1024 shaders), but running at 925 MHz with a slightly improved 81 GB/s memory bandwidth to the same 4GB of GDDR5. It also has the same TDP of 35W.

Now, I suspect that the small performance drops are due to throttling because since the two MBPs are identical except for a few faster components in the 2017 model, the hypothesis is that the second one throttles more. Why do I think this? Well, the core clock is increased 75 MHz with no known increase in cooling efficiency. Thus, the 560 will run at higher clocks more of the time, causing heat to go up, and eventually, throttling to ensue quicker. The benchmarks take a non-zero time to complete; thus, the 560 likely has to lower its clock for more of the test, resulting in a slightly worse score.

*Under Windows, using experimental modified drivers, overclocking IS possible on both models (to some extent). When using MSI Afterburner, I was able to get a stable overclock by increasing the 460's clock to 560 level, dragging the memory clock slider all the way to the right (!), and undervolting by 40 mV. The card throttles about the same amount, but performances is up on the 560 by about 5%. Alas, it doesn't persist through reboots, but that's OK for me.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.