NP9262 9800gtx sli benchmarks

I would recommend Dox since it's not a custom driver, it's simply an inf mod to allow valid NVidia drivers to run in your notebook.

http://www.laptopvideo2go.com/forum/...howtopic=19675

You only need the inf file from here. You can get the actual driver package from anywhere but you might as well dl it here also.

The score is pretty well what I expected. About on par with a desktop GTX 260 which is very good. You should have no trouble running just about any games that come out this year on high settings.

I wonder how much better the 9800's are compared to the 8800's. You have 2.83 quad cpu and 3dm6 will give a lot of points for the cpu, so i wonder how much of that 14k score is related to that bad ass cpu instead of the gpu? an ideas?

The 3DMark01, 03, 06 scores are almost entirely GPU driven and your CPU has only a very small effect on the overall score. They are 3D graphics synthetic benchmarks and they're designed to directly measure the performance of graphics cards, not CPU. I'm not too familiar with Vantage but it obviously has a separate test for CPU comparisons.

The 9800GTX should show about roughly 15-20% over the 8800GTX give or take depending on the application. SLI scaling varies greatly from one application to the other and driver selection is critical as well as in game settings. Some applications can actually have reduced performance with SLI while some are nearly double. The best way to test this is to run the same test with SLI and in single card mode to compare the difference.

Woody, I don't know if your point is entirely correct. If you take a look at the xps m1730 scores, the guys with the extreme cpu's are hitting around 14000, the others like me with the 2.4 cpu's are hitting around 10000. Everything else is equal but the cpu's, that would tell me that the 3dm06 is heavily cpu dependent.
I just want to add that these were the scores right out of the boxes.

Edit:
I'm not trying to say that he doesn't have a kick ass rig, I just want to know how good those gpu's are since I've heard they are not betther than the 8800's.

If you take a look at the following bench http://www.notebookforums.com/thread217245.html , Me and Abadee have exactly the same spec except for the cpu's. He is close to 15000 in 3dm06 (which is incredible) and I am closer to 11000. Ofcourse, I am not the only one like this, all scores sow the same outcome.

The 8800M GTX is the same basic GPU as the 9800M GT. Same core, same shader processors, same frequencies, same memory. Performance should be pretty much identical. The 9800M GTX is the same except for more shader processors and double the memory and therefore will run shader intensive 3D games faster especially at higher resolutions.

In the comparison you mentioned above the 8800M GTX in SLI are overclocked significantly and therefore not an accurate comparison to a stock 9800M GTX configuration. Overclocking the graphics cards has a large effect on 3DMark scores.

The CPU will have an effect but not nearly the amount you suspect. Just try downclocking your CPU a little and run a test. On my desktop I can run my E6750 at stock 2.67GHz or clocked up to 3.7GHz and the 3DMark06 score is only slightly different. The quad core will have a greater effect than the frequency typically but still not that much.

I can score around 10,200 with a single 9800M GT (same as a 8800M GTX) in Windows XP running a E8600 at 3.16GHz so with two of them you should be getting much better than that even with a 2.4GHz CPU. I suspect you might have driver issues or SLI issues that are affecting your score. Vista will cost you about 300-500 points typically depending on the drivers you're using.

I hope youre right woody... I will do some tweaking and get back to you. I can tell you that I have OC'd the hell out of my cards and only get about 150 more points, i thought that was odd too.

That is odd. There must be something else going on. I can't tell you how many times I've run these benchmarks and experimented. You have me intrigued though and I want to go find out just how much of an effect the CPU has on cards in this range.

I ran some 3DMark06 tests on my desktop out of curiosity and the results actually show a greater disparity than I had thought resulting from different CPU speeds.

My desktop consists of a E6750 on an MSI P35 motherboard with a slightly overclocked GTX 260 Graphics card (192 Shader units) running at 620/1080. DDR2 was set to 5-5-5-15 in all cases with a 1:1 divider.

I did three runs under Vista and one under XP. Both using NVidia driver version 177.92. All tests are rounded to the nearest 50 points for easy comparison. Default settings and 1280x1024 resolution.

Vista SP1:

• 2.28GHz (6x380MHz) 10400
• 2.70GHz (6x450MHz) 12000
• 3.60GHz (8x450MHz) 14600

XP SP3:

• 3.60GHz (8x450MHz) 15100

Clearly there is a significant difference in 3DMark06 score in this case. Keep in mind though I'm running at 1280x1024 so I'm more CPU limited with such a powerful graphics card. If I were to run the same test at 1920x1200 the results would be grouped much tighter as the GPU becomes more limiting. Last time I tested CPU performance in 3DMark06 I was using a much slower graphics card so the results were much closer together.

It is also possible to tweak the graphics control panel manually to affect the score so comparisons you read online must be taken with a grain of salt.

In the real world with the notebook graphics cards you're describing you are typically configuring your games to run at a high resolution and be GPU limited in which case the speed of the CPU is having a much lower impact on your framerates as they are in my comparison above.

Nehalem is due out late this month and should make all the current high end Core 2 Quads obsolete overnight so prices should drop like rocks. I don't know how this will affect mobile processors. Probably not much at first but we will see. I'm hoping to get my hands on a cheap Q9650 or 9770 when this happens. We will see.

I'm gald you did those test woody.... i was starting to think i was nuts, but you confirmed exactly what i was talking about. There for, everyone running the test on the free version of 06 are on a level playing feild.

I ran a couple test today also, stock clocks and sli enabled i scored 10007. sli disabled i scored 5010. then with sli enabled ang oc'd to 650/950 i only scored 10301. I did this test with dell stock drivers and the new 180.43. there was only a couple point difference between the two on all test.

One thing i totaly dont get though, laptops built with a single 8800gtx scores the same as the sli's. but disable sli and I ony get half the score. this is confusing the crap out of me..... any ideas??????

I really don't know. It doesn't make sense because your score seems very low. Are you certain they are in fact 8800GTX cards? Run GPUz (get it at cpuid.org) and confirm the settings.

Just out of curiosity to see how the number of cores scale in 3DMark06 I ran the test again on my desktop with Windows XP at 3.6GHz but this time on only a single core. I scored 11900 compared to 15100 on two cores. I don't think it's a linear extrapolation to four cores but clearly even at that clock speed the number of cores will have an effect. It depends on how well 3DMark will scale to four cores from two. Only someone with a working quad core can test that directly.

I DLd gpuz and i definately have the 8800gxt G92M, REV A2 .

It doesn't make sense to me either. How can I get 5000 on 3dm06 with one card is not right, but i am not the only one with these results.

It also seems weird that SLI exactly doubles your score. Do you have Crysis or any other games that have a standardized repeatable benchmark?

I think the only difference between your 8800M GTX and my 9800M GT is a BIOS flash.

Are you certain the system is going to full speed during the benchmarks? Nothing is causing your GPU or CPU to be downclocked? Have you verified that the test is running on both cores?

I have almost the same system specs, except the HD. I have 2 - 320 GB 7200RPM in Raid 0... Drivers are all that comes with the Sager CD.

3dMark01 - 39,415
3dMark06 - 15,970

havent tried any benchmark tool...

Just a question... do you get a BSoD if you try to connect an external display while on SLI? cozz i do and i have to reset my system before it display it to the external display.

Anyone here have a sager with HD in a raid 0... if so what is your boot time?

hey question for you 9262 owners. How loud is the laptop with 2x9800GTX's? and does a mad amount of heat come out of it?

its not that loud. well actually mine is always on max (Fn + 1). But without hitting Fn + 1 2 - 9800MGTX in SLI is not that loud, they rarely go to maximum fan speed even though im playing Crysis on high settings.

For the temp, mine doesnt give off the same amount of heat for the two. The 1st gtx is cooler than the 2nd gtx, i think its like 5F cooler.

MY temps are

idle
1st Gtx = 35F
2nd Gtx = 38F

Crysis at high

1st gtx = 67F
2nd gtx = 72F

my room temp is around 68F - 78F

hope that helps

I'm sure it's a typo but your GPU temps are in C while the room temp is in F.

yeah my bad! good looking out!

Im using HWMonitor for my temps which is in Celsius

and Fahrenheit for my room temps

would you mind posting the benchmarks, here?