CPU thottled while encoding in clamshell mode.

[ChunkAhoy]

I have done my homework and completed the benchmarks to see the performance difference between open and closed clamshell mode on my MacBook Pro 15" 2011 2.2Ghz.

I followed the exact instructions on this macrumors thread:
http://forums.macrumors.com/showthread.php?p=12184995#post12184995

I am using Handbrake with a .txt log for each encode. The result on witch i rely is the average frames per seconds that were encoded on a specific video. The average frames per seconds is directly related to the total time taken to encode a certain video.

So.... encoding the exact same video from the thread above, i get the following results:
Opened Clamshell = 37.82 fps
Closed Clamshell = 30.25 fps

This is a 20% performance loss in closed clamshell mode if you compare to opened clamshell.

It is important to note that both tests were started with the MacBook Pro's temperature stabilized with idle load.
This video lasts only 10 minutes and is encoded on a MacBook Pro 2011 2.2Ghz in around 5 - 7 minutes. Since the computer takes a certain time to reach peak temperature, the performance loss should be higher on a full length movie.

If you also have a MacBook Pro, i'd like to see the difference while encoding with closed clamshell and opened clamshell.

Regards.

 

[hehe299792458]

that's quite disappointing... but I suppose it might be necessary for cooling purposes.

 

[ChunkAhoy]

that's quite disappointing... but I suppose it might be necessary for cooling purposes.

Yes it is disappointing. I wonder if there's any way to know if the CPU is being throttled event when the lid is open?

 

[iMacDragon]

Yes it is disappointing. I wonder if there's any way to know if the CPU is being throttled event when the lid is open?

Run it for long enough in tortures mode ( both 3d and cpu maxed ) and it does, in my testing on windows, without more airflow, not much can do about it really, the CPU simply runs hotter than any previous used, when maxed.

Also remember that technically, until it's throttled to below the specced speed, it isn't actually being officially throttled as such due to the boost feature overclocking it whenever it feels able to do so. Though in my testing with the above torture, running furmark + prime95, it did severely throttle under stock speed eventually, basically, the cpu constantly adjusts speed as high as it feels safe to keep going at with thermal situation, so speed constantly varies up and down a bit.