M860ETU QX9300 + GTX280m

Can't wait to see your results.

 

I'll definitely keep you posted on that one. I'm hoping that it'll run great from the start but I have the feeling that I'll have to play around with memory timings. 3.2GHz Quad-Core is simply amazing (in a laptop) and future proof, so we'll do what it takes to make this mod 24/7 stable. Can't wait to see how much I'm going to score in 3DMark06, 14k would be so nice but impossible, I'm aiming at 13.5k or so. There's not much I can do in this laptop to push performance any further, GTX280m is OCed to the max (a few games are not even 100% stable already after an hour or so playing), and the QX9300 pushed at 3.2GHz is the limit!

 

I will also be upgrading to a GTX280M once I receive it but I will be staying at stock 2.5 Ghz CPU clocks lol. I have to say, I have been gaming on my laptop for hours (GTA IV + SR2) and the temps of my QX9300, both idle and on load are slightly lower than when I had the Q9000 installed. HWMonitor shows a 32C CPU(s) idle temp but Rivatuner shows 36C sometimes, so I am not sure what the deal is with that one. What ever CPU temp HWMonitor shows, Rivatuner usually shows 4-5C higher.

 

Oh sorry, I remember you mentioning this in your previous PM. For temperature monitoring, nothing is more accurate than RealTemp for your CPU:
http://www.techpowerup.com/realtemp/
Report back and tell us your idle and load temperatures. But I wouldn't be suprised if your QX9300 still ran cooler than the Q9000, you probably reapplied some better thermal paste and not 20 pounds of it unlike what manufacturers usually do.

 

Thanks for the info, I am trying out RealTemp right now and my idle temps are 34 - 35C. I have been playing GTA IV for quite some time now so will note the temps under load soon.

Edit1: So far maximum temps under load are 73C but I am sure they will get higher after over 3 hours of gaming. According to RealTemp, the maximum temperature of my GPU was 77C and the idle GPU temp was 38C. RealTemp is giving the same readouts as HWMonitor so it looks like Rivatuner is possibly innacurate.

EDIT2: The maximum CPU temps under load after playing GTA IV for many hours are 76C on 2 Cores, 73C on one Core and 74C on the other. My original OEM Q9000 used to reach up to 80-82C on each Core. Perhaps that was down to the thermal paste being used.

When I install the GTX 280M, I will be using MX2 thermal paste so it will be interesting to see if their will be any temp differences or improvements. It did look like there was too much thermal paste on the CPU (and GPU) when I removed the original one though.

 

Actually, for GPU temp monitoring, GPU-z is the best. I'm not too fond of all in one software that are supposed to read temps of everything.

 

I use GPU-z as well and it is good to see that it is also reporting the same temp for the GPU that RealTemp is reporting.

 

HWMonitor is much more accurate for me than Realtemp/Coretemp. they are 5c higher than my actual temperature as read by nearly any other program, including intels own programs. Realtemp is also measuring my clockspeeds poorly. am at full load and its jumping from 2.3ghz to 2.5ghz to 2.8ghz to 3.0ghz, when in actuality its at a stable 3ghz

 

Isn't that mainly due to what the Tmax temp is used by the program. That is not always clearly defined by Intel.

Either way, glad to see you guys overclocking successfully. Granted when I go the QX9300 route I will most likely keep it at 2.5GHz unless I really need it faster. Most likely I will undervolt it so I can O/C the GTX 260m.

 

I don't rely on Realtemp to read my frequency, there's CPU-z for that. But I've always used Realtemp, 5c higher is not much but I would rather read 5c higher than real temps than lower. I usually use HWMonitor to read other temp sensors on the motherboard.

htwingnut, it was for CoreTemp that the TJMax was off 10c for 45nm CPUs. CoreTemp is always used for 65nm CPUs and RealTemp for 45nm.

 

First results: extremely unstable.
The RAM indeed now runs at 1333MHz but with the same 7-7-7-24 timings, so the computer ends up freezing all the time, BSOD etc. So I'm going to start attempting to reprogram the timings.

 

Upcoming required mod: overvolting CPU because after all, the CPU also plays a role in my stability issues, it's very probably not stable at 3.15GHz at stock voltage and needs an extra notch or two on the voltage. I'll be trying a copper mod soon.

 

Seems like RJtech changed the "look" a bit.
Now I have no idea which 280M I need.

Is it the one compatible with D900F, MXM3.0, Type B
Or the one compatible with M570TU MXM-IV, MXM2.1

I tend do believe it's the later but I wouldn't bet. I haven't played around mu M860TU for a while now...

Thanks

 

You're only compatible with the MXM 2.1 variant.

 

Thanks.

Awkward thing is, that RJtech, lists MXM-IV when it should be III

I took a look at eurocom and I believe it's a typo...
http://web.eurocom.com/EC/ec_model_config1(1,198,0)
http://web.eurocom.com/EC/ec_model_config1(1,196,0)

 

Yeah you're stuck with MXM 2.1. It's pretty safe to assume that the upcoming cards won't even run on 2.1 so GPU upgradability on this laptop stops at the GTX280m. However there's some good potential regarding the CPU if you overclock a bit.

 

I'm really worried about the MXM 2.1 type. According to here: http://www.mxm-upgrade.com/Tech_13.html

Type IV is too tall to fit in my rig...

Could you guys please confirm your part number on the GTX280M?
Was it the M570TU one?

 

Yes it's the M570TU one.

 

The last MXM 2.1 card to be produced in the type IV size was the 7950M GTX. Eurocom and RJTech are incorrect in this case; the MXM 2.1 GTX 280M is a type III card, not a type IV.

http://forum.notebookreview.com/showpost.php?p=5335978&postcount=3
http://forum.notebookreview.com/showpost.php?p=5160658&postcount=5