M11X slower on battery?

I read in a review that only with mains power connected the M11X runs full speed. Especially the 335M. Is this true?

 

Yes, it is true.
In order to reduce battery consumption, the GT335m is throttled.

 

Is this something that can be adjusted?

 

I wasn't sure about this myself, so I did a quick 'generic' test.

Theory is that 3DMark06 is mostly about your graphics card.
So I've pulled up an old 'with' mains power run done in the past and did a quick 'battery only' run.

M11xR1 SU7300 OC'd 1.6Ghz BIOS A05 GT335M 260.63 OC'd

M11xR1 Mains power GPU OC'd 590/1416/912

3DMark Score 6924.0 3DMarks
SM2.0 Score 3200.0
HDR/SM3.0 Score 3389.0
CPU Score 1456.0
Game Score N/A
GT1 - Return To Proxycon 26.78 FPS
GT2 - Firefly Forest 26.56 FPS
CPU1 - Red Valley 0.46 FPS
CPU2 - Red Valley 0.74 FPS
HDR1 - Canyon Flight 33.82 FPS
HDR2 - Deep Freeze 33.96 FPS​

M11xR1 battery only GPU Oc'd 580/1392/900

3DMark Score 6916.0 3DMarks
SM2.0 Score 3220.0
HDR/SM3.0 Score 3357.0
CPU Score 1457.0
Game Score N/A
GT1 - Return To Proxycon 27.47 FPS
GT2 - Firefly Forest 26.2 FPS
CPU1 - Red Valley 0.46 FPS
CPU2 - Red Valley 0.74 FPS
HDR1 - Canyon Flight 33.06 FPS
HDR2 - Deep Freeze 34.07 FPS​

Now if games in general loosely follow the 3DMark06 results, you may be hard pressed to tell if you're on mains power or battery.

In the past, with my M11xR1 I've been hammering away on the game Crysis with an HDMI to a Full-HD monitor and cause I was still using the M11xR1 keyboard, I happened to notice the power light was orange and not blue. Doh! power cable had slipped out.

If it wasn't for the orange power button light, I wouldn't have even noticed.

So in summary, if the GT335M is throttled when on battery, I can't tell the difference.

 

The 335m is NOT throttled on the battery. Maybe the CPU is, but not the GPU.

The battery should last about 1:45 at full blast, while gaming.

 

As we see the CPU Score is unchanged. It may be the case that due to the oc of the cpu the throtteling under battery is disabled. What do you think?

 

Very good results - but I can give a clear poor man's easy proof of the power question on the R-1.

I play Champions-Online heavily; it is CPU dependent; with speedstep off and the system plugged in, CPU oc'd at 1.6 I get say 18 fps in a certain test area I use even with full power options - if I plug it in literally the fps jumps like 5 or 6 frames. Unplug it - then right back down.

 

Start> control panel> power options> change plan settings> change advanced power settings> Graphics processor> on batter = max performance.

My computer used to stroke out for a few frames when on battery. this fixed is. The GPU will be sett on "balanced mode" when on battery, you just have to change it manually (even if you select high performance mode)

 

I checked and this was the problem apparently; thanks!

+1 Rep

 

ejohnson is correct.

But be aware that this will absolutely kill battery life. It is the trade-off you make for higher performance on battery.

 

I think its a good trade off... I would rather loose some battery life and be able to play the game smooth with no random drop off.

 

Thanks, ejohnson!

 

Edit - Sorry was doing SQL and didn't note the focus had changed.
So, anyway I'd be so bold to say, on the correct power setting there is no difference between mains power and battery on an M11xR1.

We can probably ignore the rest below.


Bit odd then, that on my M11xR1 the 3DMark CPU score and the the two CPU graphic intense tests are the same on both battery and mains power.

I guess I could always query a 50,000 record SQL table in some complex manner or on an un-indexed field with mains power off and mains power on.

Query results are logged in milliseconds and only use CPU and memory, with no regard about GPU.

So here they are:

Queried a bit in a SET data-type (Not indexable) using MySQL on about 56,000 records and with the mains power on it took 0.156sec.

Resetting the query cache and doing it again with only battery power and I get exactly the same result 0.156sec.

So, in summary, if I can get identical result on an SQL query using 56,000 records with or without mains power... then the environment the MySQL server is running in must be the same.... yes???

Maybe the reviewer that mention the throttling on the M11x under battery usage was referring to Planck time (tP) and that it failed their scientific benchmark.

Guess one could always run wPrime 32M with and without mains power and you'll get your own results to the nearest second as it's not as precise as a Database engine.

Mine were: 58.813sec with mains power and 58.033sec without mains power.

 

It's throttled even when not on battery, but only when idling or when it gets too hot (IIRC it gets downclocked to 185Mhz).

 

even if the energy option for GPU is set to max, it only eats battery if the gpu is set to nvidia, right?