The Best Gaming Video Cards for the Money: February 2007

[pschmid]

Detailed video card specifications and reviews are great - that is, if you have the time to do the research. But at the end of the day, what a gamer needs is the best video card within a certain budget.

 

[Datman]

*The Gainward BLISS+ 7800 GS cards, which actually have 7900 GT and 7900GS GPUs, are also fast AGP cards on par with the X1950 PRO. These are far more expensive and can only be purchased in Europe, however.

And Australia. :twisted: :twisted: :twisted:

 

[cleeve]

What, isn't Australia part of Europe?

Kidding!

 

[sojrner]

the hierarchy chart is great... have not seen someone put something like that together in... well, ever really. So many of us "know" where these cards stack but never actually look at it in terms of tiers like that. Great job Cleeve.

 

[cleeve]

Thanks for the Kudos, Sojorner.

 

[torque79]

there's already one out there, but I second the kudos. that hierarchy is BRILLIANT. I will be very interested to see how it changes when the next ATI line comes out.

 

[blade85]

I second the kudos. that hierarchy is BRILLIANT. I will be very interested to see how it changes when the next ATI line comes out.

took the words out of my mouth . Nicely done

 

[Featherstone]

The hiearchy is cool... but wrong...

it shows the 7950 GX2 higher than the x1950xt and the x1900xt. You can clearly see that in almost every benchmark (every one i looked at on thg vga charts) the x1900xt beats the 7950 GX2 Let alone the x1950xt and the x1950xtx!

(the x1950xtx KILLS the 7950 GX2... by like 80% @ high res/af/aa)
(78fps compared to 50fps in HL2, 61 compared to 48 in Oblivion etc.)

I looked at the hierarchy for about 7 seconds and saw this mistake, which seems to me a VERY major one. I hope that it is one of the few. I feel bad for the poor sap who spends $500 on a GX2 and then gets outperformed by a $250 x1950xt...

 

[bourgeoisdude]

the hierarchy chart is great... have not seen someone put something like that together in... well, ever really. So many of us "know" where these cards stack but never actually look at it in terms of tiers like that. Great job Cleeve.

I tried to do something like a hierchy before, but it was just a mess. For one, the chart wasn't exactly right just due to the fact that I did not know every card in between and such. The hierchy is super awesome, no doubt about it. Keep it up.

One thing I wonder these days--not that I'm incredibly interested or anything--but what about mobile video chipsets? Not like your busy enough already spending hours of your time finding out the best video cards for the money or anything (I'm teasing of course), but will tom's ever publish any comparisons of mobile gpu benchmarks? There's like practically 20 different GeForce 7 series mobile graphics cards, and mobility radeons, Intel GMA 950s, etc...again you are doing awesome with your stuff and you could use a break, but maybe you could convince another Tom's Hardware dude/dudette to come up with a comparison/benchmark/hierchy chart of the mobile graphics chipsets?

One more thing Cleeve--IF you rated the Intel GMA 900 or GMA 950 in the hierchy (I realise they are not gaming chipsets here, just wondering where they stand as far as performance), where would they be? Intel, SiS, and VIA/S3 are certainly not gaming chipsets by any means, but do any of those even compare to, say, the FX 5200? If these graphics chipsets aren't your fortè, then ignore my questions--I mean, who besides me even gives a rats?

Still--I'm curious, so if you happen to know, I'd love to hear from you. Otherwise, ignore my fruitless curiousity...

 

[cleeve]

Please provide some benches to back that up dude.

The high res benches with the GX2 I've seen shows it whopping the X1950 XTX.

 

[Featherstone]

PLease provide some benches to back that up dude.

The high res benches with the GX2 I've seen shows it whopping the X1950 XTX.

I did...
I quoted THG VGA charts.

(the x1950xtx KILLS the 7950 GX2... by like 80% @ high res/af/aa)
(78fps compared to 50fps in HL2, 61 compared to 48 in Oblivion etc.)

Here is a link if you want.
http://www23.tomshardware.com/graphics.html?modelx=33&model1=517&model2=586&chart=211

 

[MayDay94]

Great job, cleeve. i always enjoyed your posts and now it's nice to see it as a full fledged article. i also liked the performance benchamrks on the older amd64 chip in other article.

 

[cleeve]

Here's the difinitive GX2 vs X1950 XTX review at Xbitlabs:

http://www.xbitlabs.com/articles/video/display/quality_vs_quantity_19.html

The summary:

Half Life 2:
GX2 wins at 1280 res and higher

Oblivion:
X1900 XTX wins, but it's close. With FSAA the X1950 XTX pulls ahead

Battlefield 2:
GX2 wins

COD 2:
GX2 wins

Far Cry:
GX2 wins (close though)

FEAR:
GX2 wins

PREY:
GX2 wins

Quake4:
GX2 Wins

Serious Sam 2:
GX2 wins (close though)

Hit Man: BLood Money
GX2 Wins

Splinter Cell: Chaos Theory:
GX2 wins (close though)

Tomb Raider Legend:
GX2 Wins

Titan Quest:
X1950 XTX Wins

IL2:
GX2 Wins

Age of Empires:
X1950 XTX Wins

Rise of Nations:
GX2 Wins

GX2: 13
X1950 XTX: 3

 

[sojrner]

honestly, since the last rev of the vga charts many here have called the sli setup (which includes the gx2) into question. When you look at any other bench you see the gx2 just as you describe cleeve... but here at tom's the vga charts paint a false picture. Any way of finding the issue there and resolving it or what?

 

[Featherstone]

Thats good to know that xbit has benchies that better represent true performance (i think). but what was the THG article based off of, if not their own VGA charts? kinda odd

A. that their VGA charts are wrong
B. that their article was apparently based on other peoples benchmarks?

 

[Synergy6]

Go for Xbit over THG when it comes to the "definitive" article on graphics cards?
Synergy6

 

[Featherstone]

Was that a question? If so, I would have to simply say, i dunno. But xbit's benches fit this THG article better than THG benches fit.. their own article...

 

[Slobogob]

I love the GPU charts comparing the different Radeons and Geforce cards. It´s pretty brave to publish it since every dimwit will question it (based on their subjective observations). Great article!

 

[bourgeoisdude]

Comeon guys, let's not get Cleeve in trouble here

 

[sojrner]

meh, using other benchies as a base for an article is not bad in-and-of itself... but having resident charts that contradict said article calls things into question. Like I said, since the last rev of the charts most ppl here knew there was something up with the sli marks, but n00bs will really be scratching their heads.

I only asked a question. 8)

Don't want you gettin in trouble there Cleeve. 8O Still think these are great articles man.

 

[Featherstone]

meh, using other benchies as a base for an article is not bad in-and-of itself... but having resident charts that contradict said article calls things into question. Like I said, since the last rev of the charts most ppl here knew there was something up with the sli marks, but n00bs will really be scratching their heads.

I only asked a question. 8)

Don't want you gettin in trouble there Cleeve. 8O Still think these are great articles man.

I agree completely.

I don't think that questioning why an article from a site and benchies from the same site don't agree is exactly "dimwitted"...

 

[torque79]

even if THG's benches did not have SLI issues, they never test cards with as many games as quoted by cleeve above (16!!!). Obviously for this kind of article, it is necessary to consider a site with SO MANY MORE games benchmarked than THG ever does.

 

[Featherstone]

even if THG's benches did not have SLI issues, they never test cards with as many games as quoted by cleeve above (16!!!). Obviously for this kind of article, it is necessary to consider a site with SO MANY MORE games benchmarked than THG ever does.

I agree... My sole point is that THG benches and the THG article do not agree and that is strange.

 

[cleeve]

but what was the THG article based off of, if not their own VGA charts? kinda odd

A. that their VGA charts are wrong
B. that their article was apparently based on other peoples benchmarks?

Don't worry about me getting in trouble fellows! There's no conspiracy or subversion here, nothing that I'm worried about the world seeing anyway.

I'm not commenting directly on THG's charts, which are certainly useful as a resource.

However, when I make a recommendation or write an article I take all sources of information into account, not just one site's... THG or not.

I used Xbitlab's review as a representation of the norm, but it's not like they're my only source of info either. If you look across the web you will see that most GX2 vs X1950 XTX benches agree with it. Of course, certain benches like the ones at THG do not jive perfectly with it, but there are always variables in hardware setups and testing methods.

So, without having personally reviewed a GX2, my method is to look at all available sources of info when forming conclusions. Which is what I've done here, and I stand behind that.

As a writer for THG, I don't have a problem with saying that I respect Xbitlabs or any number of other hardware sites. I consider them colleagues, not rivals.
There's alot of people doing alot of good work out there, I would never plagiarize it but I don't have a problem with citing a respected source as justification for my conclusions.

As far as suggesting the article is 'based on other people's benchmarks', that's pure sensationalism. The bulk of the article is about the best cards for the money. The Hiearchy chart is a small part of the article, and it's based upon my experience and knowledge. Yes, some of that knowledge comes from thoroughly reading up on all the material I can get in addition to personal experience, but this isn't a benchmark comparisaon - it's a general hiearchy chart.

 

[jordanwise]

tell you what, i wouldn't mind benchmarking games for a living like all these people at these sites seem to do, they're living the dream