Are these benchmark closely correct

found this on Factorgaming.

 

Compare to:
http://www.videocardbenchmark.net/gpu_list.php
Click on your video card it displays a bar graph.

 

passmark? pfft... i'd stick with the graph posted by icon007, looks about accurate to me (although 8800m-gtx is a bit low - i've gotten over 10k with my setup).

 

i dont know where they get them scores from.
the 9800gt and 8800gtx are basicly the same card and the gts is roughly 5% lower power (% i think) so how the gts scores highest i have no idea.

 

The 4850 remains to be seen, but will probably rank higher with the right drivers.

 

so wheres the 9800gts in this list.
i can see the 8800gts but not a 9800gts

 

9800M GTS scores higher than the 8800M GTX and 9800M GT? No.

9700M GTS scores higher than 8800M GTS? No.

3850 Crossfire is not > 8800 GTX SLI

The list is rubbish without CPU and driver listings anyway.

 

hmm, here are the cpu with gpu benchmarks

 

i hope this helps

 

not really
they are still misleading scores.
above post shows 9800gts = 9683
9800 gt = 9416
8800gtx = 9194
whoever releases them stats must be doing it for their own gain.

if you really want to see real 3dmark06 scores look no further than the actual users themselves HERE

dont forget there are various scores here with base scores and o/c scores.

 

Well...8600M GT is definetly wrong.

DDR2 scores around 4500
DDR3 scores around 5600

 

This list is garbage. Disregard its validity. It sucks. Et cetera.

DDR2 can't hit 4500 unless it's heavily overclocked, and tested at 1280x800.

 

8600m GT is sub 4000 @ 1280 x 800 unless it's with a 2.4GHz+ proc. And an extra 200MHz on the memory clock won't get you 1100 extra points in 3Dmark06... You must be a dell rep Akuma

 

1280 x 800 and 1280 x 1024 screws with a lot of results. Even some 9800 laptops have 1280 x 800.. like that gimped best buy asus

 

Kidding?
My previous laptop (Vostro 1500) had DDR2 version, and with low overclocking I got around 4500 .. (T7250 2.0Ghz, 2GB ram)

 

Indeed, 9800M GT > 9800M GTS.
How can they even make an objective benchmark with so many graphic cards? Its obvious they didn't use the same laptop somewhere.

I would prefer to rely on this one:


Source: Notebookcheck

 

Really?? I trust you... But now i trust no-one else

What scores on stock clocks?

Thanks

Oh and notebookcheck for the WIN

 

My previous laptop was also a Vostro 1500, with T7250 and 3GB ram.



My highest recorded score, at 650/?/525, 1280x800. The score probably won't be much over 4k at 1280x1024.

 

650 core is quite an oc It would explain the 4600 since shaders are linked to the core.

 

I just remembered that the shaders must've been 1300, because I kept them linked with the core in Rivatuner.

 

thats very heavy oc. 8700mgt has a core clock of 625.

 

What resolution? 1024x768?

8600M GT DDR2 will barely pass the 4k mark @1280x1024, and that's at a very heavy OC [600++ core/500++ mem]. It'll get to 4500 at the same OC when tested @1280x800 though.

DDR3 will easily pass 5k mark at even the slightest OC, of course @1280x800.

@Topic: the graph is a bit conservative [in the sense that the scores are at the lower limits of possible results from the cards, given the various CPU's that can be matched with them] but yeah it's kind of accurate. Although I wonder which laptops come with ATI MR 4850?

 

why the hell is everyone bringing up results done at 1280x800, they are completely useless, the standard is 1280x1024, if your laptop doesnt support it get an external. the graph is actually quite accurate and probably the best you will get because its not as easy to make a comparable 3dmark graph like with desktop systems.

 

The graph is not accurate for ranking mobile GPUs, by score or real world performance. The 9800M GT, 8800M GTX, and FX3600M will all outperform the 9800M GTS; the 9600M GT will not outperform the 8700M GT; the 9700M GTS is inferior to the 8800M GTS. Neither Mobility X2 will outdo the 8800/9800 in SLI.

Test all of the cards on the same CPU with the same drivers, then we'll see the real truth. When I want a compiled list of random scores, I go to Notebookcheck.

 

Yeah, us with the 1440x900's will just do that for your benchmarking happiness.

 

You really have to watch the resolution with 3DMark06. Anything benchmarked under the standard resolution of 1280x1024 is invalid in my book. You need to benchmark at 1280x1024 if you want to be able to compare to other systems.

As harsh as this statement may be, I fully agree with the poster.

 

if you have a 1440x900 res screen do it at 1440x900 then, there is only about a 3% difference between that and 1280x1024, dont forget to mention you did it at 1440x900 though.

^ at Kevin_Jack2.0, so there is a few that are a bit off, the question was if they're closely correct and they are. it would be impossible to do what you say as well, because not all mobile gpus are MXM-II/MXM-III compatible and they are very rare even as well.

 

Just a note, seeing as to how the FX3600M is based on the 8800M GTS core, I don't see why it should outperform the 9800M GTS.

My 8800M GTX SLI scores around 12k with stock clocks and a E8300.

 

Based is a key word, the FX 770M is "based" on the same core as the 9600M GT, and beats it out. While most 9600Ms score in the 5200-5400 range, the FX 770M scores 5800-6000+.

 

The problem with this is that, the 9800M GTS is also built on the same core, but with substantially higher clocks.

At any rate, my 8800M GTS scores higher than that with a T9300.

Without the CPU information, one should not attach any importance to the ranking or scores of the cards listed.

 

My mistake; I meant to include it after the 9700M GTS but copy pasted into the wrong sentence.

The list is incorrect enough to be disregarded as a reliable tool for comparison. Surely you won't argue against this fact. It can't be considered "closely correct" at all.