Hi guys. This is my first post on the forums and i know this issue was covered many times here but my situation is a bit different. I've posted this to the apple discussion boards too but nobody answered
Me and my GF were playing ARMA 3 (bootcamp) some time before and I noticed my rMBP performing WAY worse than her classic 15 inch macbook pro. Went to the game's config menu to see the FPS: mine was fluctuating between 19-26 fps while her was steady at 40fps. First I thought that may be an issue about the game or windows installation so i switched our macs to OS X investigate further.
I've made our macbooks render the same 3D scene at the same time, and started monitoring the heats on iStat Menus. Her CPU was going between 95-102 Celcius while mine only fluctuated between 85-92 and dropped to 80's after a minutes and slowed down on rendering. I've opened up console and saw:
16/07/14 12:06:00,000 kernel[0]: IOPPF: Sent cpu-plimit-notification last value 1 (rounded time weighted average 1)
16/07/14 12:06:00,000 kernel[0]: IOPPF: Sent gpu-external-plimit-notification last value 0 (rounded time weighted average 0)
That may be a legitimate thermal throttling but I realized i have some weird problem when we booted back into Win.
Opened the same scene in our games and I started monitoring the clock/power/temp/voltage values in AIDA64. Her machine never downclocked, temps stayed at GPU:76 | CPU cores: ~90-100C. But my GPU never hit over 69, CPU never hit above 75 and there was SEVERE downclocking going on in CPU clock.
I was never able to make out the culprit in this problem because if one of these 4 values go down, the other 3 will go down too. I was never able to make out which one is the initiator. May be any of these 4 or maybe something I didn't even think of.
I tried locking the CPU multiplier with throttlestop and this made the GPU downclock after a while. I tried locking the GPU clocks with NvidiaInspector, but that did not help too. Finding a good balance between these two makes the game a bit playable. If I don't lock any clock values, first the CPU clock drops, then the GPU starts to throttle ( GPU is not always the case. it happens from time to time. But CPU can't stand 4 minutes of gameplay)
I've took the mac to the applestore and my applecare was just over 3 days ago back then. The genius told me the dust accumulation might have damaged the heatsink or maybe even the logicboard, and the repairing cost might vary between 200$ and 1700$ (sorry but i'm not paying that money for a machine with a bottomcase which I can't open for cleaning the dust off)
To make things even clearer:
we have same power settings in Win/Mac
we are playing games with the identical configs
we rendered the 3d scenes with identical rendering modules(which utilizes CPU)
our macs' hardware is almost identical (same ram,gpu,vram and CPU)
I tried resetting the SMC/PRAM several times. never worked.
I've tried turning power nap off.
this is the thread I've opened in apple disc:
https://discussions.apple.com/message/26307008
I thank all of you that reads this long post and appreciate any help that you may give to me. It pisses me off to pay for a chipset that doesn't work fully and being held responsible for dust accumulation inside a sealed laptop case
Me and my GF were playing ARMA 3 (bootcamp) some time before and I noticed my rMBP performing WAY worse than her classic 15 inch macbook pro. Went to the game's config menu to see the FPS: mine was fluctuating between 19-26 fps while her was steady at 40fps. First I thought that may be an issue about the game or windows installation so i switched our macs to OS X investigate further.
I've made our macbooks render the same 3D scene at the same time, and started monitoring the heats on iStat Menus. Her CPU was going between 95-102 Celcius while mine only fluctuated between 85-92 and dropped to 80's after a minutes and slowed down on rendering. I've opened up console and saw:
16/07/14 12:06:00,000 kernel[0]: IOPPF: Sent cpu-plimit-notification last value 1 (rounded time weighted average 1)
16/07/14 12:06:00,000 kernel[0]: IOPPF: Sent gpu-external-plimit-notification last value 0 (rounded time weighted average 0)
That may be a legitimate thermal throttling but I realized i have some weird problem when we booted back into Win.
Opened the same scene in our games and I started monitoring the clock/power/temp/voltage values in AIDA64. Her machine never downclocked, temps stayed at GPU:76 | CPU cores: ~90-100C. But my GPU never hit over 69, CPU never hit above 75 and there was SEVERE downclocking going on in CPU clock.
I was never able to make out the culprit in this problem because if one of these 4 values go down, the other 3 will go down too. I was never able to make out which one is the initiator. May be any of these 4 or maybe something I didn't even think of.
I tried locking the CPU multiplier with throttlestop and this made the GPU downclock after a while. I tried locking the GPU clocks with NvidiaInspector, but that did not help too. Finding a good balance between these two makes the game a bit playable. If I don't lock any clock values, first the CPU clock drops, then the GPU starts to throttle ( GPU is not always the case. it happens from time to time. But CPU can't stand 4 minutes of gameplay)
I've took the mac to the applestore and my applecare was just over 3 days ago back then. The genius told me the dust accumulation might have damaged the heatsink or maybe even the logicboard, and the repairing cost might vary between 200$ and 1700$ (sorry but i'm not paying that money for a machine with a bottomcase which I can't open for cleaning the dust off)
To make things even clearer:
we have same power settings in Win/Mac
we are playing games with the identical configs
we rendered the 3d scenes with identical rendering modules(which utilizes CPU)
our macs' hardware is almost identical (same ram,gpu,vram and CPU)
I tried resetting the SMC/PRAM several times. never worked.
I've tried turning power nap off.
this is the thread I've opened in apple disc:
https://discussions.apple.com/message/26307008
I thank all of you that reads this long post and appreciate any help that you may give to me. It pisses me off to pay for a chipset that doesn't work fully and being held responsible for dust accumulation inside a sealed laptop case
Last edited: