Question:
5672WLMi Tests
Ok, I'm starting this tread for some tests/benchmarks that people did with their 5672 WLMi, specifically for the 1 gig, 100 GB version.
This is my first laptop ever, so let me know of any mistakes i do, i shoulda studied harder in my electronics classes!
FutureMark05
First test I did was FutureMark05. with "out of the box" specs, meaning FAT32 and no omega drivers. I got a score of 1602, or was it 1620? either way, low 1600's.
MobileMeter
Next, I did a little with MobileMeter. during my 24 hour burn-in, the temperature fluctuates from 46-52 degrees Celsius. the fan will come on at around 51. It takes maybe 20-30 sec. to go from 46 to 52 degrees, which I thought was really fast considering the computer was just idle. The HDD went to aprox 43 degrees. My friend playing WoW logged 46 degrees. MobileMeter was used when I did other tests.
Battery Eater Pro
After reading about the USB/battery problem with WinXP and the dual core chipset, I decided to do my own testing. So far, only 2 tests were done with BEP in the Classic mode. This consists of loading the CPU to 100%, and the GPU with a OpenGL app @ 800X600. for both tests, MobileMeter was loaded. I found out I couldn't load the battery to the max, because the LCD screen would automatically lower the brightness to the lowest level after a couple minutes of inactivity. there is aprox. 15W of difference between the lowest brightness and the highest.
With a mini ubs hdd drive (1inch) from kayser plugged in, I had a discharge rate of aprox 26W. When the LCD brightness is maxed out, discharge rate jumps to 40W, and temperature quickly rises to 59 degrees. when the brightness is lowered again, the discharge rate returns to aprox 26W, and temperature immediately falls back to normal idle level. This lead me to believe that the temperature graph in MobileMeter is not for the CPU, but for the system or the system heat sink. I don't see why the CPU would get hot when the LCD brightness is increased. with the USB device plugged in, I got a battery life of 147 mins.
WithOUT the USB device attached, I had a discharge rate of approx 20-22W. With the LCD on Max, the discharge rate jumped to 36W. so the USB device sucks aprox 4W as indicated by MobileMeter. The system temperature also falls immediately when the LCD brightness is lowered. without the USB device, I got a battery life of 174 mins. 27 mins more battery life.
These times roughly match when a calculation is done. My battery is 14.8V, 4800mah. 14.8 X 4.8 = 71.04 Watt hours, altho BEP says a full charge is only 64watt hours. but sticking to the calculation this means it can provide 71 Watts for 1 hour. or 35.5 Watts for 2 hrs. or 23 watts for 3 hours.
I'll post my SuperPi test scores when i get the chance.
Upcoming Test: undervolting...the touchpad on this thing is too hot for my tastes. my bro's ThinkPad is ice cold compared to this. first i need to figure out how..
Answer:
Re: 5672WLMi Tests
I just ran SuperPi and Prime95 to get both cores loaded up 100%, the LCD is up at 100%. CPU load is taking from taskmanager.
at 100/100/100 (100% CPU load) the discharge rate from MobileMeter is aprox 46Watts
at 0/100/100 (SuperPi stopped, aprox 50% CPU load)) the discharge rate is aprox. 39 Watts.
at 0/0/100 (both SuperPi and Prime95 stopped, aprox 1% CPU load) the discharge rate is 26 Watts.
at 0/0/0 (even at 0% brightness the screen is still very viewable) the discharge rate is at 22 Watts
I think the reason there is a jump of watts from idle to 50% CPU load is because in taskmanager it shows as Prime95 using both cores, but only to 50% each.
I think i read somewhere, that one core could not enter a deep sleep state if the other one was working.
sleep time for me.
Answer:
Re: 5672WLMi Tests
nice man...
Thanks for this. Oh, and please also post some detailed gaming scores as and when you get the opportunity! :-)
Answer:
Re: 5672WLMi Tests
Aw geez!
I just loaded NHC, and none of the voltage controls options were available. Maybe because I'm also running the Acer Empowering software.
So I loaded up RM Clock, and it correctly detected the Core Duo, but under the PM Features, it only has "N/A".
I was thinking undervolting a dual core would be a little wierd. Would I need to have both cores loaded up to check voltages? would i be able to control each core individually?
Looks like neither RM Clock or NHC can do anything with the voltages. I hope its just that I have "factory settings" with all the Acer power management software installed thats stopping me from undervolting.
Answer:
Re: 5672WLMi Tests
Undervolting
I got my hands on RM Clock 2.05 Pre-Release yesterday, and have been playing with it ever since.
My 5672 is a little on the hot side for my tastes, and all this talk about undervolting in the HP/Compaq forum got me going. So, I downloaded the pre-release of RM Clock that now supports Core Duo and Core Solo.
I used RM Clock in conjunction with MobileMeter, SuperPi and Prime95. SuperPi and Prime95 were used to load the CPU up to 100% (both cores). MobileMeter was used to see how much heat was generated and the discharge rate of the battery.
With the tests that I did, I was able to undervolt the dual core T2300 down to 0.965V at 10.0X, down from 1.263V which is a difference of 0.298V at 10.0x. With the new lowered voltage, the max temperature dropped from 73-74 degrees celcius down to 59-60 degrees celcius!! the touch pad is less hot at the max load(could be either the GPU or the CPU in that area). Prime95 did not report any errors or warnings.
The discharge rate of the battery has also fallen. I used to be able to reach a discharge rate of over 50W (LCD and CPU maxed). This would have given me just a little over 1 hour of battery life at Max load to the CPU. Now, I can stretch the battery to almost as if I was running at the lowest clock setting!
HOWEVER, Intel in their smartness disabled the ability to undervolt at 6.0x. So the minimum voltage is still stuck at 0.950V. I guess they really want people to buy their Low Voltage and Ultra Low Voltage CPUs, afterall i'm sure they must have spent LOTS of money making them.
So in conclusion undervolting this bad boy will not only reduce heat, but will also give you better battery life, altho Max Battery life cannot be increased because voltages at 6.0x cannot be lowered past default.
Answer:
Re: 5672WLMi Tests
Very interesting indeed! I would love to try this out, as I agree, the touchpad simply gets excessively hot.
Nice work.
Edit: just snagged a copy...this is going to be fun ;)
Answer:
Re: 5672WLMi Tests
hey...i think maybe making a guide to how you did this will help not only me but many others..your the man
Answer:
Re: 5672WLMi Tests
Well, I've found my "stable" voltage at 10X to be a bit higher than yours, gino_lee. Total system crash at 0.975, total stability at 1.000V, so that's where mine is staying ;)
Of course, if Acer in their infinite wisdom hadn't put the RAM almost directly under the touchpad we might not have to worry so much...:rolleyes: Haven't even found a cooler pad that "catches" that area sufficiently to be worth it. If anyone thinks that I am over-reacting, the ambient temps here in Malaysia are 30-35C with 90% humidity. Heat hurts! :(
Recommend anyone who hasn't checked out the forums at rightmark to do so, but if gino_lee fancies doing an undervolting FAQ as suggested, then that would be cool.
Edit: FWIW, and FYI my normal use CPU temps range from 53-65C, and my HD temps from 45-53C. Scorching. HD runs at 50-51 average, which I don't like. Might have to create some more ventilation....
Using Mobilemeter, with the voltage reduction discussed, dropped from 74C, to 65C. Thank you for the headsup on this.
Answer:
Re: 5672WLMi Tests
If you are having heat issues, you can also try a little trick that i found works really great. I never thought it would work until I actually tried it after reading in HP/Compaq forum about the extended battery sticking out of the unit.
All u need to do is just put a pad of paper or something at the back of the laptop to prop it up. My laptop has about 1inch inclination, and you'll be surprised using MobileMeter that the temperature drops 5-6 degrees within a minute.
NB: the amount of temp drop from this little trick depends A LOT on ambien air temperature.
Maybe i'll write a VERY simple guide for the RM Clock tonight. Barebones version
Answer:
Re: 5672WLMi Tests
If you're REALLY in a pinch for excessive heat issues (and it is and issue with this laptop), then go into the ATI control Center, and pick the "Advanced View". Then go to POWERPLAY. Make sure that the setting is set to "Optimal Battery Life" when on batteries. and when Plugged-in set it to "Balanced" or even to "Optimal Battery Life".
Changing these setting will clock the GPU core, and RAM down to about 100/100, from about 400/400 (Numbers from NHC). This will reduce heat and increase battery life.
Honestly, 90% of laptop users (me included) won't be pushing the X1400 to its limits. Even at the "Balanced" level, I can watch a high quality 1.4gig divx movie.
Only problem with that is that for some reason the control center will sometimes change the settings on you for no apparent reason. So, I'm just waiting for ATI Tool to support the X1400, last time i checked it was only the X1300 and X1600. :(