Waiting for a GTA V benchmark!
Runs 40+ fps at 1680x1050, default settings. Definitely very playable and looks very good. The benchmark tool averages at 30fps but again, the in-game performance is much better.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Thoughts on AMD Radeon R9 M370X over the Nvidia 750m?
- Thread starter MartinAppleGuy
- Start date
- Sort by reaction score
Runs 40+ fps at 1680x1050, default settings. Definitely very playable and looks very good. The benchmark tool averages at 30fps but again, the in-game performance is much better.
Its a fun idea, I don't think it will happen though. But just to dream even further, lets imagine they put an upgradeable slot in the back of the monitor where you could throw in any standard graphics card.
Its a fun idea, I don't think it will happen though. But just to dream even further, lets imagine they put an upgradeable slot in the back of the monitor where you could throw in any standard graphics card.
I was thinking the same: Apple Cinema Display 5k with optional GPU.
Sounds like a better upgrade than a lot of us thought. Keep the game experiences and benchmarks coming
Oh man I hope that's true.
Because I was really not expecting that and the thought of importing my Witcher 2 save (that was started from a Witcher 1 save) makes me all kinds of giddy
Hi Guys,
I just upgraded from a 2012 MacBook Pro to a 2015 one and did some quick GPU benchmarks. I'm not much of a gamer but this new machine is very nice (SSD is seriously fast). The GPU is quicker than the 650m
Results:
3dmark Version 1.1 - Windows 8
1 - MacBook Pro Retina 2015 15" i7 2.5GHz, 16GB Ram, 512GB SSD, R9 m370x
2 - MacBook Pro Retina 2012 15" i7 2.6GHz, 16GB Ram, 256GB SSD, GTX 650m
Cloud Gate
1 - http://www.3dmark.com/cg/2868874 - 11581
1 - http://www.3dmark.com/cg/785007 - 9798
Fire Strike
1 - http://www.3dmark.com/fs/5289695 - 2269
2 - http://www.3dmark.com/fs/864635 - 1411
Ice Storm
1 - http://www.3dmark.com/is/3055020 - 88723
2 - http://www.3dmark.com/is/880967 62350
I just upgraded from a 2012 MacBook Pro to a 2015 one and did some quick GPU benchmarks. I'm not much of a gamer but this new machine is very nice (SSD is seriously fast). The GPU is quicker than the 650m
Results:
3dmark Version 1.1 - Windows 8
1 - MacBook Pro Retina 2015 15" i7 2.5GHz, 16GB Ram, 512GB SSD, R9 m370x
2 - MacBook Pro Retina 2012 15" i7 2.6GHz, 16GB Ram, 256GB SSD, GTX 650m
Cloud Gate
1 - http://www.3dmark.com/cg/2868874 - 11581
1 - http://www.3dmark.com/cg/785007 - 9798
Fire Strike
1 - http://www.3dmark.com/fs/5289695 - 2269
2 - http://www.3dmark.com/fs/864635 - 1411
Ice Storm
1 - http://www.3dmark.com/is/3055020 - 88723
2 - http://www.3dmark.com/is/880967 62350
Probably not if Apple kept the size the same after they removed the DVD drive from the cMBP. I'd take a 970m over a thin MBP with a weak GPU.
Last edited:
That's because Apple's cooling system is so bloody weak. I mean, come on. One puny little heat pipe for both the CPU AND the GPU? Come one. At least put 2 heat pipes on them, there's plenty of room for that.
One reason: just one heat pipe.
Look at this. This is the inside of a Razer 14, with a design similar to the rMBP (one can say it's a 1-1 copy but anyway, heck the Razer is 0.01" thinner than the MacBook Pro). Yet it is able to pack in a GTX970M while Apple can only put in a M370X and even that thermal throttles. Why is that? Well, the Razer's cooling solution is waaaay better. Double heatpipes for both the CPU and GPU while all of them are connected to the GPU in one way or the other so actually 4 heatpipes on the GPU which is located left in the picture. Compare that with the rMBP which only has one tiny heat pipe running across both the GPU and the CPU. As soon as something GPU intensive has to be done it heats up the CPU as well, or vice versa. The GPU pushes out too much heat for that one heat pipe to handle which oh by the way is actually also cooling the onboard GPU on the quad core i7. So actually one tiny heat pipe is cooling a CPU and both the dGPU and the iGPU, all at once.
This is why the rMBP's cooling system is weak. In fact most if not all of the Macs Apple has to offer thermal throttle at some point. iMacs doing anything CPU intensive will thermal throttle. MacBook Airs: same story except it's now one heat pipe cooling both a CPU and a GPU. Those ****ers get hot when put under load. MacBooks: same story, except this time it's more a question of why someone is doing CPU intensive work on a MacBook but anyway. Same goes for the Mac Mini.
The only Mac that cools well is the Mac Pro because it has one massive heatsink with a large fan to suck air trough.
It is possible for Apple to put a 970M in the macbook pro. However, this would consume more power and would be louder than the macbook pro already is. No matter the cooling solution, this is true. Apple has always errored on the side of conservative, cool, quiet, better battery life vs packing as much power/heat as possible into a confined space.
If you aren't happy with this trade off, then there are many gaming notebooks available to choose from.
If you aren't happy with this trade off, then there are many gaming notebooks available to choose from.
How's it speculation? It's a fact Apple can do a much better job at cooling their components, even if they're not planning on beefing up the hardware. Hell, the current mid-range hardware is thermal throttling which is nowhere near as power hungry as a 970M.
Woah. Easy fella, it seems to me you're waaay over exaggerating there. If you check out iFixit's teardown of the 15" rMBP with dedicated dGPU it's in the same ballpark as the picture you posted. It has a single heatpipe across both the CPU and GPU which in total to my eye looks about the same size as all four of the Razer's. Have I measured both? No. But neither have you.
The Razer might be a bit better (I have no idea), but you're using a bit too much hyperbole to be taken seriously.
Last edited:
One reason: just one heat pipe.
Look at this. This is the inside of a Razer 14, with a design similar to the rMBP (one can say it's a 1-1 copy but anyway, heck the Razer is 0.01" thinner than the MacBook Pro). Yet it is able to pack in a GTX970M while Apple can only put in a M370X and even that thermal throttles. Why is that? Well, the Razer's cooling solution is waaaay better. Double heatpipes for both the CPU and GPU while all of them are connected to the GPU in one way or the other so actually 4 heatpipes on the GPU which is located left in the picture. Compare that with the rMBP which only has one tiny heat pipe running across both the GPU and the CPU. As soon as something GPU intensive has to be done it heats up the CPU as well, or vice versa. The GPU pushes out too much heat for that one heat pipe to handle which oh by the way is actually also cooling the onboard GPU on the quad core i7. So actually one tiny heat pipe is cooling a CPU and both the dGPU and the iGPU, all at once.
This is why the rMBP's cooling system is weak. In fact most if not all of the Macs Apple has to offer thermal throttle at some point. iMacs doing anything CPU intensive will thermal throttle. MacBook Airs: same story except it's now one heat pipe cooling both a CPU and a GPU. Those ****ers get hot when put under load. MacBooks: same story, except this time it's more a question of why someone is doing CPU intensive work on a MacBook but anyway. Same goes for the Mac Mini.
The only Mac that cools well is the Mac Pro because it has one massive heatsink with a large fan to suck air trough.
What is problematic about this analysis is that the rMBP was found in all cases to power throttle (the wattage on the power brick is too low) before thermal throttling. I don't know the intentions of other users who state that what you've mentioned is not fact - but I can tell you that the rMBP's processor runs between 70-80 degrees Celsius under a full GPU and CPU load like prime95 and furmark being run simultaneously, this is in a stable state after 15 minutes of running. The cooling system on the rMBP is superior to the majority of laptops given its thin dimensions, Apple has just ensured that the cooling system is never tested near its maximum cooling capacity by giving us a power constrained laptop.
My rMBP reaches 99C in no time. Under load it is absolutely impossible to keep 70-80C temps on the CPU. Only the 750M stays around there. It needs about 35W of the CPU to go to 6000rpm and reaches 99C and then it stays between 95-99C for the duration of any load. That is the temps the cooling system and the clock throttle targets. 70-80C you must have a very different machine. If it does not hit the 99C limit it will just clock faster until it is that hot.
The Asus UX501 keeps its temps at 85C but it has a much more powerful cooling system than the rMBP. Two heat pipes, thicker heat sink and peak noise is much much lower (somethin like 4000rpm) while still not throtteling as soon as the rMBP. rMBP also has much lower base cpu clock speed which should be because it has the GPU but with the dGPU active it does not really go above its base speed either. There is just less heat to handle on either front (cpu and gpu) and it barely keeps up.
I really think Apple should limit fan speeds to 4500rpm max and beef up the heatpipe heatsink part of the equation. It is just not good enough as it is, or at least could be a lot better. More heatpipes and deeper heatsink would not require any additional case thickness but could greatly increase cooling area and heat removal.
The Asus UX501 keeps its temps at 85C but it has a much more powerful cooling system than the rMBP. Two heat pipes, thicker heat sink and peak noise is much much lower (somethin like 4000rpm) while still not throtteling as soon as the rMBP. rMBP also has much lower base cpu clock speed which should be because it has the GPU but with the dGPU active it does not really go above its base speed either. There is just less heat to handle on either front (cpu and gpu) and it barely keeps up.
I really think Apple should limit fan speeds to 4500rpm max and beef up the heatpipe heatsink part of the equation. It is just not good enough as it is, or at least could be a lot better. More heatpipes and deeper heatsink would not require any additional case thickness but could greatly increase cooling area and heat removal.
Strange, I found my 2011 MBP would start to cannibalise the battery, i.e. under heavy load the battery would disacharge despite it being plugged in. I used to get this with Starcraft II.
Can't comment on the newer models, I've not had one.
It's by design, all MacBook Pros do it, including the very earliest 2006 Core Duo MBP. If you ran that machine without the battery it limited the CPU to 1GHz to ensure it could provide sufficient power.
I don't think that was what fat jez meant. I also have a 2011 Macbook Pro (non removable battery) and when my computer is below 50% mark of battery and the CPU is being subjected to heavy load, the computer stops charging and even starts to DISCHARGE despite being plugged in.
I've only had one other Macbook before this from early 2008, the black one with 2,4 GHz Core 2 Duo and I strongly remember that laptop being able to charge despite being low on battery and the CPU being subjected to heavy load.
I have not yet tested this with any rMBP 15", maybe the cause would be the quad core CPU being so powerhungry. Anyway, if my battery charge level is over 50% then it doesn't discharge during full CPU load which is a little strange...
The computer can demand more power than the AC adaptor alone can provide, so it dips into the battery. It's by design and is normal behaviour for any MacBook.
I've had a 2006, 2008, 2011 and the current rMBP.
I've had a 2006, 2008, 2011 and the current rMBP.
Last edited:
Just read through the entire thread, was hoping to read more about throttling issues. I've benched my m370x with 3dmark and found that it doesn't throttle nearly as much as when I'm playing GTA V. I've turned the CPU way down to see if that helps but doesn't. The m370x not only throttles but peaks at much lower MHz in GTA V than in 3dmark. Makes me think they did that on purpose for benchmark numbers. The laptop runs GTA V great for about the first 5 mins and then you can see the micro stuttering from the GPU throttling. Still playable but irritating. Would be great to stop the GPU from throttling because it runs relatively cool, just above 60c and doesn't draw too much wattage, around 70w peaking to 91w. Are others having similar experience?