M17x Benchmarks thread !!!

Yes. All my posted scores are with the 1333Mhz Samsung 2x2GB sticks but I have the 1066Mhz sticks in right now. I'll switch them first.

 

the RAM thread was moved as it was more specific to the RAM verses the unit itself.
thus a side topic.

 

You're the boss!

 

The Xfastest drivers don't work. They give us the same results from the other drivers we tried.

Alitunay described it best:

 

I seem to have more driver crashes. Am I supposed to run nhancer with stock settings or just lax the OC clocks?

 

it's just an easier program to use then nvidia system tools. it syncs perfectly with it. I am testing it and memset to tell you the best settings for both. right now messing with memset and the timings aren't doing anything like the whitebook.


I will report back with what I find and you should do with both.

 

I brought my GPU OC settings down and getting stable results. Now I'm starting to creep it up again.

 

Wait so what are you guys using instead of Ntune.... so far I like Ntune, but is there a better way (ATI tools, capable of reading 280m's?) or do you have something else?

 

rivatuner was the best for me with my 8600m gs but it's not reading the new drivers. it has so many tweaks it it.

 

Do NOT use nTune for OCing! It's ridiculously unstable. Wait a week or two, then get the version of Precision that supports the 280s.

 

there are others like rivatuner or the coolbit hack but i couldn't get either to work properly it seemed to me the nvidia system tools was the only one i was able to get to work.

 

I think I'll probably just stick with Ntun... I am able to control virtually everything... mind I have to do it all by hand.... mouse.... but it's working out well for me so far

 

I'm using the nvidia system tools.

 

that is ntune, just packaged into what nvidia now calls their system tools.

 

I figured that

 

SF4-No AA. I'll post both my settings and results.

 

dammn mayne! where ya been at? hope all is well with you out there.

 

messed with nhancer and memset and my stock scores from 10135 to 10362. I wil do more testing to see what's the best setting from it.

 

which sli mode are you running with nhancer?

 

I stll have one gpu. the second one isn't coming until 2 weeks. but afr should be the best one. you can try them all.

I don't have that option since I hveone card but the ati verion is afr. alternative frame rendering. where one gpu does a frame then the next gpu does the next and so on.

edit 10362 to 10382. I maybe at it's limits but i'm going to see what else I can use to push it stock then overclock and then I can post what I did.

 

so I'm understanding 1:1 is seemingly impossible with both 1066 & 1333

 

Hoping there some magic wand that will do it but it seems unless we can bump up the FSB and have it stable it's going to be hard. I would want to know if any other laptops are achieving it.

 

hmm, should definitely be getting [much] higher, could be drivers?

 

I have nothing to compare it with. Did you run it?

 

Maybe Fullscreen: Off is giving him a hit on FPS. Lots of games run worse in windowed mode.

 

I think it was just off to take a screen shot.

 

Get the free trial of fraps... it can take picitures and 30 sec video...

 

Then something's up because Vicious's W90 was getting 126 FPS with 8xAA

http://forum.notebookreview.com/showpost.php?p=4959986&postcount=5

 

I tried using nvidia system tools to improve performance but it's making the screen unreadable. Still trying to tweak.

 

Thanks. Gives me an idea of what the numbers should look like.

 

You can see it in game. Someone else with a M17x needs to run it.

 

You have vsync activated, so your frame rate won't go over 60. My current and busted Dell XPS M1730 with a single 8800 scores well over 80 average with AA. Try the test without vsync.

 

can you provide a screen shot?

 

Not at the moment, because that'll involve running the benchmark again, and my GPU health is practically done for as I am running the machine in safe mode. But I can assure you that you should be getting much more with your card; just disable vsync and try the test again. BTW: Street Fighter 4 benchmark does not use SLi as it operates in single-gpu mode. The game engine is very well optimised and not demanding at all.

Edit:
I just found a text log of my last performance.

 

I wasn't questioning vsync I just wanted to see your other settings. Can you let me know what they are?

Does it support Crossifre because people with the ATI cards are doing much better.

 

Thanks.

10 char.

 

No prob. It's the best info I could provide on my part at this time. FYI: I amped up all the settings to their maximum.

As for the crossfire support. I have no idea. Normally, SLi is the more mainstream and supported form of dual-GPU solution for most games, but that general norm doesn't mean that this benchmark tool won't be an exception that supports crossfire. It just might.

 

Vsync is off for me and AA settings just bring the score down. That leaves changing drivers but I'm not willing to do it just for this demo. On to the next one.

 

That's very strange. And yeah, it is probably your drivers. 182.xx series are very stable, but not the fastest. I was using the 186.03 drivers; seem to give me boosts in games using DX10, as well as the DMC 4 benchmark and of course, SF4 benches.

This thread has shed light on a lot of questions that I had about the 'All Powerful'; I look forward to seeing more benchmark results.

 

It seems 185.85 has the right SLI profile for SF4 otherwise rename the exe to Crysis.exe thats what i have gathered thus far.

 

Renaming it to Crysis did bring the average up to 100 w/8xAA.

Edit: I bet the Quads do much better in this benchmark.

 

Been reading the forum for some time now and can't find it if it is here. Anyone have benchmarks or any type of scores for a system with q9000 and sli 280 gtx? My unit is still in production at the moment. I have another question, do you all feel like you made a wise purchase?

I am looking forward to my new laptop, but is it worth it. I know beauty is in the eye of the beholder; just want to get some opinions.

 

There has been limited participation in this thread from M17x owners. Waiting for more people to get their hands on them. I've seen a T9600, QX9300 and P8600 benchmarks here besides my own.

There's a thread in the side topics section answering your other question.

 

Here is my desktop score for comparisons

SCORE: 10005
AVERAGE: 136.19FPS
OS: Windows Vista(TM) Home Premium
CPU: Intel(R) Core(TM) i7 CPU 920 @ 3.70GHz
Memory: 6134MB
Graphics Card: NVIDIA GeForce GTX 260x2
Display Setting: 1920x1080 60Hz 8xAA

Im running 186.08 which has the SF4 benchmark profile for SLI also.

Not sure it even used SLI though so im going to run it again with the SLI meter on.


Ran it again and same score i also renamed the Exe to Crysis as well. The SLI visual indicator was turned on and its clearly using SLI .


Seems that anything over 2xAA might be using the DX10.1 path which would explain the jump in peformance on the W90 . Still need to do some more digging .

 

Hey quad, what do you get with no AA on?

 

Thanks MAndrake for the info about the side topics section. You can expect to see my test results whenever I receive my unit.

 

Just ran it for you , 248fps with no AA on. So 112fps loss with 8xAA on LOL.

 

Ouch. This seems like a trend with nVidia and capcom games. Nonetheless, it's hardly anything as the difference in fps isn't noticeable at all when playing the actual game.

 

Yes ive noticed this trend with Capcom as well but your right seeing as anything over 60fps is fantastic . Ill try Devil may cry 4 and Lost planet when i get the new system and see what i come up with plus Resident Evil 4 PC is coming soon.

I thought though Alitunay gotter better scores then anyone to date with Devil may cry 4 DX10 and 8xAA on ? So this may just be a driver thing im not really sure and i need to load some new drivers on my desktop anyhow. What is the newest driver the M17x can run without issue ?

 

186.24 are out... in some hour's ther will be a modded inf for 280m ... be aware this can be a good jump in performance...