What benchmark results matter to you?

what matters with benchmarking?

  • 3dmark, aquamark, etc.

    Votes: 9 5.1%

  • game benchmarks

    Votes: 105 59.3%

  • both

    Votes: 47 26.6%

  • none(don't care)

    Votes: 16 9.0%
  • Total voters
    177
Although game benchies are the most important, I like to see how they perform in 3Dmark. Simply because it give me an idea of how well it handles DirectX and how well the CPU handles multithreading.
 
My 3dMarks are only important to me for stock, to make sure I am on par with other systems to see if I am still running solid.

Game benchmarks for every other feasible reason regarding a gaming comp.
 
Game benchmarks..

I only really look for the games that I play, and then everything else is second to that.
 
Talonz said:
Game benchmarks..

I only really look for the games that I play, and then everything else is second to that.
Click to expand...
Exactly. Of the "new" intensive games out their, I only play Counter Strike Source. So I only care how HL2 benches right now. I look at the system they compare it with... and then the res they play it... and how many FPS it gets... then I can judge off their.
 
Here is some inside info for ya.

The guy who runs the 3DMark05 thread for this forum (me) uses both 3DMark05 and game benchies!

I just really love organizing data and 3DMark05 gives me plenty of opportunity to do that.
I think 3DMark's most important use is to determine if your computer (3DMark05 focuses
a lot on the video card though) is functioning correctly and optimally.
 
I'll use anything that's good as a benchmark.
Then as a benchmark I will seek out the best performance I can get.
Always compareing my data to my benchmark.
Other then that they are pretty much useless
 
games ofcourse, but like to see just what its capable of in 3dmark. doesnt matter if it only gets 1 3dmark as long as it gets great fps in games.,
 
/start rant

games duh!!! lol

hehehe screw 3dmark...it's a engine that is not even used in ANY game..and gives false info of what games can run

/end rant
 
benchmarks are a bunch of shit. as long as the games I play are fine, I DON'T GIVE A SHIT
 
I think games provide a better real world appraisal of what a card is capable of, from an end user perspective.

I like to see the addition of games such as Everquest 2 which showcase graphics quality as much as FPS.

More and more reviews seem to be capped off with a 'best playable settings' table lately, which is a feature that truly has more significance than almost any other single page of a review.
 
Siciliano said:
My 3dMarks are only important to me for stock, to make sure I am on par with other systems to see if I am still running solid.

Game benchmarks for every other feasible reason regarding a gaming comp.
Click to expand...

Ditto- benchmarks also help to see if a tweak yielded a gain. I know it's not a gaming benchmark, but the score helps measuring relative performance.
 
Both types of tests have their place. Synthetic tests do allow one to compare parts of complex GPUs more fairly compared with an asymmetric comparison (e.g. running a game). The overall experience can't be forgotten either, though.
 
AT the LEAST I must see an apples to apples compare if game benching is all that is done.

the problem with just the synthetic benchmarks is they are so manufacturer tuned... pay futuremark and get the next version tuned to your brand...

I don't know... I often wonder why there hasn't been a demand for an open source platform independant benchmark. One where everything is very well documented and accounted for... so there is no room for funny stuff... but in the end real world performance is what matters regardless.
 
Yashu said:
AT the LEAST I must see an apples to apples compare if game benching is all that is done.

the problem with just the synthetic benchmarks is they are so manufacturer tuned... pay futuremark and get the next version tuned to your brand...

I don't know... I often wonder why there hasn't been a demand for an open source platform independant benchmark. One where everything is very well documented and accounted for... so there is no room for funny stuff... but in the end real world performance is what matters regardless.
Click to expand...

Well the same can be said slightly for games to. Doom 3 and HL2 are pretty good examples.
 
Skrying said:
Well the same can be said slightly for games to. Doom 3 and HL2 are pretty good examples.
Click to expand...

Yea, but in the case that you have a game like Doom 3 or HL2 tuned for your graphics card, the end user actually benefits. How many users play 3dmark05, and how many users play Doom 3 or HL2? ;)
 
Talonz said:
Yea, but in the case that you have a game like Doom 3 or HL2 tuned for your graphics card, the end user actually benefits. How many users play 3dmark05, and how many users play Doom 3 or HL2? ;)
Click to expand...
Yes, but if a particular manufacturer's card wins every comparison employing such high-profile games only because those games were optimized for that card, then do those results mean anything when it comes to other games which might not be that biased?

At least in my view, technical specifications and performance of components of a GPU individually and as a whole often mean a lot more than using 'black boxes' in the form of games to test GPUs without regard for subtle variables.
 
I like to know my SiSoft, 3DMarks01/03/05 benchmarks purely for comparative reasons. But as for actual game play goes, as long as the game plays higher than 33fps or whatever it is where your eyes cant notice a difference, its fine with me.

Since my LCD's max res is 1280x1024, my 9800Pro is still able to hand most games at my max res.
 
Game benchmarks all the way! Most importantly, the Doom 3, Half Life 2, and UT2004 benchmarks!

(I was going to also say Quake III, but when a game gets more than 600 FPS with its settings MAXED OUT, I dont think you can really consider it a benchmark anymore :p )
 
the only reason i use 3dmark and aquamark if for comparitive reasons only. i want to be able to see if my hardware is performing the way it should be and 3dmark is a good way to tell if it is or not.
gaming benchmarks matter more to me because thats real world performance that i can see if i will be able to play that particular game at 16x12 4aa/4af or if i can play it at 16x12 no aa or af.
 
What a bunch of bullshitters in this thread. Nobody cares about 3DMark scores, yet the "post your 3Dmark score" thread has like 2000 replies.

I know it's not considered [H] correct to admit that you like the 3DMark programs, but like it or not they can be very helpful when tweaking your system.
 
Both, actually. Although the game benchmarks are much more important, synthetics can be fairly helpful at comparing cards.
 
both..I like 3dmark benchmarks to compare previous system's performance vs. the one i have now. Also to see how much of an improvement an overclock or memory setting will give me after tweaking.

But the thing that will actually make me upgrade to new components is gaming benchmarks...if they're far better than the system i'm using.
 
pigpen said:
What a bunch of bullshitters in this thread. Nobody cares about 3DMark scores, yet the "post your 3Dmark score" thread has like 2000 replies.

I know it's not considered [H] correct to admit that you like the 3DMark programs, but like it or not they can be very helpful when tweaking your system.
Click to expand...

I believe the title of the poll is "what matters with benchmarking?"

not "what matters with tweaking"
 
Both.

Generic benchmarkers are good for generally tuning your sytem. A memorytimings here, more aggressive settings there 'til you can run your benchmarkers (3DMark05, Aquamark, etc.) to the highest possible without crashing or artifacting. Then once you've got a good general setting to start with, you can fine tune it to the game engine of your choice.

In the etc. part there should have also been other applications. Benchmarking isn't just limited to the video stress. There's also sisoft sandra, PCMark and a ton of other system benchmarkers which are actually more effective than a standard game.
 
These things are ok for a bit of fun once you've built your new rig. It was fun to test mine to get a score. However, it seems some folks see them as some sort of holy grail and develop an unhealthy attitude that means they purely live their life by them and dont actually do anything else with their PCs.

Its a bit like what we call in the high-end hifi business the 'sick puppy brigade'. They have a $30000.00 hifi system but only play the same 4 CDs on it.

If my machine will run though a complete lap of 3DMark and a single pass of Memtest, its ready to go. Never had any problems with that method for 10 years or more.
 
Overall benchmarks only matter to me when they are in direct comparison to other cards that are available. Showing me that Doom 3 gets 67 fps on a certain card doesn't mean anything to me by itself. Reviews like that don't mean anything to me. Now showing me that Doom 3 gets 67fps on Card A and 52 on Card B and 78 on Card C means something. That way I can quantify it and say ok this card is this fast compared to what I have and this fast compared to the fastest card available.
It would also be good to see a benchmarking system that uses the price of the cards. Getting 120 fps is great but not so much if the card costs $700 (7800 I'm looking at you). Getting 70 fps may not be super duper but if the card is 1/4 the price of the expensive card then it seems pretty good. A kind of "fps for money spent" system would be cool to see. Sure it may not seem fair to compare the prices of a top of the line new card to cards that are a year old but if the fps aren't that much different then maybe it is ok.
Yeah, mainly I'm just mad that I can't swing the money for a 7800 right now :)
 
3D Mark is good to check the size of your e-penis or to see if a new set of drivers are worth it. But even then just because new drivers gives you a higher score does not mean that they are going to be better than the set you have. Case in point the 77.72 drivers give me a higher 3D Mark score but offer no noticable increase in the games i play (HL2, D3, WOW) at the 1280 res i play at. Also, with 77.72 drivers i get corruption if i play a game after watching a .wmv file with wmp. I think the thing that matters most is what level of image quality and framerates can you get w/ a given card and driver set. I really do not care if i get a zillion 3D Marks if I can not play a given game at decent frame rates.
 
Game benchmarks at high resolution. The main advantage I've seen come of the SLI rollout is forcing the reviewers to start using 1600x1200. Before that, they stopped at 1024x768 and called it a day, making the review useless to me.
 
the only benchmarks that actually make me see the true worth of a card are in the games that i would be playing.

personally, i really don't care about how many "marks" (both aqua- and 3d- or any others...) mostly because i don't use them. i mean, if i can play any game i want, with the visuals at something that looks nice, then thats what i want to know about.

the problem with the synthetic bench marks is that it gets down to almost a grading system. it isn't based on actual performance as far as 100000000 3dmarks doesn't necessarily mean that when you have a whole swarm of enemies firing at you, that your video card would be able to keep up with all that action. yeah it's nice to know that you can render some pretty things, some action, but you have no control over what happens, you have no involvement.

if you are able to play these benchmarks (almost as if it were a demo....) then that would mean more to me than letting it play through, almost like a movie.

basically, i want to know how a card can play through the randomness you would find in a game, instead of the same, predefined "tracks" that aquamark or any of the others would go through.

just my 2 cents.
 
Bugalaman said:
benchmarks are a bunch of shit. as long as the games I play are fine, I DON'T GIVE A SHIT
Click to expand...
WELL FUCKING SAID.

Synthetic benchmarks scores are for crybaby's measuring their e-penis with each other.

Having said that, I only care what GPU plays games best right now (in my pricerange of course), and even then, I only look at how my favorite games perform (FarCry, UT2k4 etc.), skipping benchmarks of Half-Crap, Splinter Suck etc. altogether. :)
 
I like both because if your card doesn't perform well in 3dmark05 then it won't perform well in modern games with all the graphics turned up, even though I generally rely on game benchmarks more then 3Dmark05benchies.
 
Gaming benchmarks.

I could give a rat's ass how the "latest and greatest" performs in basket weaving media arts and squirrel hunting encoding.

But some might.....

...somewhere.
 
Arklight said:
How are you going to be able to tell how the hardware will perform in your games if you don't look at benchmarks?
Click to expand...

uhhhh, by seeing how it performs in those games?

you really don't even need to look at FPS numbers, you can feel if a game is fast enough or not
 
I like and use them both. I still run Code Creatures for the hell of it every now and them. There is nothing wrong with synthetic bench marks as long as no1 is cheating.
 
I think its funny that people measure their PC with 3dMark. You should check out the people in guru3d forums when they are looking for the best drivers to boost their marks..Its saddens me that people would spend so much time trying to get 5 more 3dmarks...It's rather pathetic actually. All you need is a decent card and your all set to play games that YOU bought that card for, no need to see if it works or not..
 
Brent_Justice said:
uhhhh, by seeing how it performs in those games?

you really don't even need to look at FPS numbers, you can feel if a game is fast enough or not
Click to expand...


Don't you have to read the benchmarks in order to tell how it is going to be performing in those games?

What I mean is, that you look at the game benchmarks to get an idea of how a piece of hardware performs in games (hopefully the ones you play, lol).

It sure beats going out and buying a piece of hardware and "hoping" it runs well. lol
 
Arklight said:
Don't you have to read the benchmarks in order to tell how it is going to be performing in those games?

What I mean is, that you look at the game benchmarks to get an idea of how a piece of hardware performs in games (hopefully the ones you play, lol).

It sure beats going out and buying a piece of hardware and "hoping" it runs well.
Click to expand...

I think you are missing my point. A gamer knows if the game they are playing is being rendered fast enough for them. If it isn't they lower quality settings until it is, if it is super fast then they raise settings until it becomes to slow. They find out what in-game settings, what resolution and what AA and AF settings play best with a game. You don't even need to look at FPS numbers to do this, it's a feeling kind of thing.
 
You must log in or register to reply here.
Back
Top