Bootcamp gamers on 650m and 750m ?

[dusk007]

I never thought about this P2 methode of overclocking. I will test it out.
So far I could figure out that Apple apperantly lowered the throttle temp to 75C which is insane because you reach that so easily and those chips run fine up to 90C. I limited the CPU severly and when the chip stays below 75C its clock are stable at 925Mhz. I am almost certain that is a software issue. Still that needs to be fixed. It looks like they limit performance on purpose so they can increase lifeexpectancy of the chips. The throttle temps weren't that low before. It used to work fine on 80C.

 

[baris3]

I never thought about this P2 methode of overclocking. I will test it out.
So far I could figure out that Apple apperantly lowered the throttle temp to 75C which is insane because you reach that so easily and those chips run fine up to 90C. I limited the CPU severly and when the chip stays below 75C its clock are stable at 925Mhz. I am almost certain that is a software issue. Still that needs to be fixed. It looks like they limit performance on purpose so they can increase lifeexpectancy of the chips. The throttle temps weren't that low before. It used to work fine on 80C.

I was at the service and tech guy said :

"Yeah, there was no fail at apple diagnostic tests as you have predicted, and there is no obvious issues."

So he basically said he can not change motherboard without any defect, and when I ask him to change it because he believed me about the motherboard defect. And he said if he change the board, he will pay for that motherboard later when apple finds out the old motherboard is not defective. He told me to call apple care, and tell the issue so maybe I can get a replacement of motherboard for the sake of customer contentment. I was going to do that, but after receiving machine clocks are stable at stocks !

The tech guy ran the GPU tests for so many loops like I told him to do so. Maybe because of that, clocks does not fluctuate anymore like before. GPU diagnostic tests does not cause any heat, by the way.

After receiving machine and testing, clocks were stable like 1-2 hours of playing and went down to 800ish core for a moment, then fixed itself. Down voltage/clocks value was different this time, because memory stayed same and only core clock downclocked for a moment.

It is been two days, and there is no fluctuations at all. Temps are same as before. I don't max out fans, the things is thermals are very good with this machine. So I really don't think it is because of temps. Maybe new motherboards need some time to adjusts themselves ? I know it does not make any sense but after running a lot of diagnostic test for GPU, my 750m clocks are stable at 925/1254 now. It was not like that before service, as you know the whole story.

It may seem fixed now for mine but I can only hope for it to stay the same. I can't tell if there is power drainage from battery, yet. And I can't tell what will happen to clocks if I play for more than 2-2.5 hours since I play in sessions of 2 hours mostly.

I suggest you to give it some time with the new logic board. Maybe run GPU diagnostic test for a lot of loops ?

If you suggest any graphical-intense game from the ones you tested, I will test it and we can see if my stable clocks will go down too. I have 40 GB free on bootcamp.

 

[baris3]

I guess we can use nvidia inspector to lock things on windows...
I did it in my PC desktop with a gtx660, but I no time to test it on bootcamp (macbook pro) yet.
here:

http://www.overclock.net/t/1267918/...-disable-boost-fixed-clock-speed-undervolting

Look at the last post (from 3dwin4tor):

http://www.tomshardware.com/forum/372385-33-wanna-overclock-650m-660m

the thing is:

1- Make two .bat files to execute: bat 1 - normal speeds , bat 2: fixed or overclock speeds:


like this (made for his laptop):

C:\Users\xxx\Desktop\OC\nvidiaInspector\nvidiaInspector.exe -forcepstate:0,0 -setGpuClock:0,2,1220 -setMemoryClock:0,2,2580

we only need to modify for our macs

I'll keep looking for better understanding...

It does not work. It may work in windows PC. Apple 650m/750m cards are different, and have different BIOS.

 

[dusk007]

I still think that the throttle temp is just set too low and lower than it used to be.
The thermal paste might need some time to cure so it will likely get slightly better but the temp limits won't change. I think it can be solved with a simple software update. I could do it myself if one could edit the bios on mobile gpus. It is not faulty hardware or require any new logic board, it is just an idiotic unrealistic setting in the videobios.

 

[baris3]

I still think that the throttle temp is just set too low and lower than it used to be.
The thermal paste might need some time to cure so it will likely get slightly better but the temp limits won't change. I think it can be solved with a simple software update. I could do it myself if one could edit the bios on mobile gpus. It is not faulty hardware or require any new logic board, it is just an idiotic unrealistic setting in the videobios.

Hi,

Played Heroes of Storm yesterday, clocks were 925/1254, no fluctuations at all. I make gpu-z log files, and I could not find any downclock. Because it seems fixed with time, it does not seem to me as a BIOS/driver issue. It could be about thermalpaste, I think after some years it will go back to it's worse performance.

 

[baris3]

I still think that the throttle temp is just set too low and lower than it used to be.
The thermal paste might need some time to cure so it will likely get slightly better but the temp limits won't change. I think it can be solved with a simple software update. I could do it myself if one could edit the bios on mobile gpus. It is not faulty hardware or require any new logic board, it is just an idiotic unrealistic setting in the videobios.

The downclocking/fluctuating is back.. Card has been running stable at normal clocks no matter what temp/load is for WEEKS. Now, it fluctuates again like before under slightly heavy load. It is a temp related motherboard defect I can say, from my experience. Or wait, I should call it "very badly engineered apple motherboard", to be exact.

What is your news ?

 

[dusk007]

It is weird. I spent 2 days gaming on a lan and had no problems. Clocks stayed at 925Mhz. But aside from a little CS/Go it was mostly CSS so the whole system wasn't really taxed at all. The rest was games with even less requirements.
It is weird but I doubt sending it in would fix anything. It suspected that after the curing of the thermal paste it ran better. That it is back on yours contradicts that theory. I have decided mine works well enough.

 

[sabbyp]

i've recently noticed my 750m throttling at anything above 75c, clock goes from 1020 down to 825 and stays there. In new games like MGSV:TPP it knocks off a pretty noticeable 15-20fps or more.

Not sure what's going on but maybe the new nvidia drivers have something to do with it? might be time to roll back

 

[baris3]

It is weird. I spent 2 days gaming on a lan and had no problems. Clocks stayed at 925Mhz. But aside from a little CS/Go it was mostly CSS so the whole system wasn't really taxed at all. The rest was games with even less requirements.
It is weird but I doubt sending it in would fix anything. It suspected that after the curing of the thermal paste it ran better. That it is back on yours contradicts that theory. I have decided mine works well enough.

There is no way to fix this very badly engineered, defective macbook motherboards.

Mine ran the same game 925mhz for weeks after "cured", but now the same game fluctuates again.

Let me know if you experience any fluctuating at all. Try CS/GO with higher settings/resolution. Because it seems the problem just temproraly fixes itself for awhile, and comes to life again. At least for me.

Like before, it can be running at 925mhz for at least 10-15 minutes if macbook is cold and unused for hours.

From gpu-z logs, I read at the end of this 10-15 minutes card reaches 88 C. It mostly starts to downclock after that.

 

[baris3]

i've recently noticed my 750m throttling at anything above 75c, clock goes from 1020 down to 825 and stays there. In new games like MGSV:TPP it knocks off a pretty noticeable 15-20fps or more.

Not sure what's going on but maybe the new nvidia drivers have something to do with it? might be time to roll back

Hi,

I suggest reading the whole thread if you've not, it is pointed out as a motherboard defect. It's not driver neither anything else related. Try posting gpu-z screenshots that is taken while playing high-taxing games.

 

[ckWTB]

More powerful processors generate more heat which in turn increases the chance of throttling and reduced performance. Unfortunately, for most laptops, throttling and reduced performance is a normal way of controlled temperatures. The most direct way to address this issue is to increase the laptop’s cooling capacity with enough supplemental cooling to eliminate throttling even when overclocking the GPU.

 

[ayeying]

The downclocking/fluctuating is back.. Card has been running stable at normal clocks no matter what temp/load is for WEEKS. Now, it fluctuates again like before under slightly heavy load. It is a temp related motherboard defect I can say, from my experience. Or wait, I should call it "very badly engineered apple motherboard", to be exact.

What is your news ?

Could be dust. I had a lot of dust in my rMBP and noticed a huge reduction in gaming performance in boot camp. Almost made me buy a Windows PC. Cleaned it out and it runs like normal again