5970 vs 5870Xfire vs GTX460SLI. Results starting to come in!!

[Bandit1]

my file server, an x3 720BE, can do 4 5870s at once if i really wanna. can swap in my friends 965BE as well since i dont think he's using that comp right now, idk when i will have budget for a better cpu right now lol, was gonna get bulldozer when it came out. if you want 5870 trifire though my friend shmee just got his second and his board can do 3, and he has a gulftown . can also bench 5870 cf setups on 4ghz core 2 quad. have farcry 2, fallout 3, bf:bc2, crysis and warhead, shmee has modern warfare 2, and a few others idk what else. we have 2 900w and a 1050w psu between us as well, so we can load up some pretty power heavy systems if we really want

Very interested in 720be results,although mine is unlocked to x4.I would think it will hold back some...

 

[Keysplayr]

Very interested in 720be results,although mine is unlocked to x4.I would think it will hold back some...

I have the same. Well, it's in my sons computer. It unlocks to X4, but didn't bother keeping it that way.

 

[bryanW1995]

replied to pm... I'm in

2x reference 5870's
I7 930 (haven't really finalized my oc yet, just got my TRUE)
ocz 6gb triple channel kit
pc power and cooling 750
3007 HC

I have

Crysis
Warhead
Batman AA- Spoof id will eqaulize my AA numbers to nVidia counterparts
Bioshock 2 (in the mail)
STLAKER clear sky
Mass Effect 2
Call of Duty 4, WAW, MW2
GHostbuters
Far Cry 2

And lots of other less demanding games

make sure you go push/pull with the true. I had trouble going over 3.8 with just the "pull" fan, now I'm at 4.2 stable after adding the "push".

I don't find tests conducted by NVidia focus group members useful...hence my comment.
Perhaps we could ask an ATi rep to do the same and see how closely the results correlate.
Keys can post all his results first(yes Keys,first) followed by the results conducted via an "alternative view".
I'm sure the results would be interesting as focus group members are well known for presenting "just the facts Mam".....aren't they?

well, considering that keys and apoppin did something similar several years ago and he has an impeccable testing reputation, I'm inclined to believe his results. What have YOU done other than troll? Why don't you offer to help with the testing?

by the way, sli kicks the shit out of crossfire right now, he doesn't need to screw with the results to get a positive (for nvida) result.

 

[Elganja]

have we figured out the first game and how (more like where in the game) we are going to test?

 

[Keysplayr]

Maybe we should try this. Would everyone please post when you are available for chat and for testing. I'm available most evenings EST after 9pm.
Keys

 

[Elganja]

anyday, but NOT today, this week after 9pm works for me

 

[Keysplayr]

anyday, but NOT today, this week after 9pm works for me

Tuesday or Wednesday after 9pm is good for me. Anybody else?

 

[Sho'Nuff]

Maybe we should try this. Would everyone please post when you are available for chat and for testing. I'm available most evenings EST after 9pm.
Keys

Ditto, but I am not available this weekend (Fri-Sun) as I will be out of town for a wedding.

 

[mrblotto]

Sorry, my SLI 8800GTX's seem too old. I'd be like the fat kid that gets picked last for kickball.....lol. Sounds fun tho!

 

[dug777]

If you run any demos I'd be happy to throw in a few maverick oc'd 5850 results, but my newer games library is pretty light-on at the moment (currently working through some Serious Sam HD, Tales of Monkey Island, FEAR and some casual NFS:HP2 and Colin McRae 2005

 

[Annisman*]

Has to be after 11PM for me. Unless on Thurs, have the day off.

 

[Keysplayr]

Guys. Apparently there has been a huge uprising about this thread stating that the only reason I started it is because I know the outcome and would show Nvidia in a good light.
What they don't really know, is that I DONT know the outcome. My primary interest in this topic and said video cards listed in the thread title, was due to the two latest H articles where everyone is complaining that the 10.7 driver were broken and the article was totally inaccurate. One guy even asked why I didn't include 5830 Xfire in my original post because that would be the direct price performance competitor for the SLI'd 460s.
This member claims to belong to something called the anti-shilling coalition. If he wants to represent such a cause, maybe he would better represent it by just asking me first instead of mile long accusational posts when he doesn't even really know anything.
If he bothered to read the rest of the posts in this thread, he would have seen that all other benchmarks are welcome and will be added. Didn't bother. Just saw my thread title and went nuts. This kind of behavior needs to stop. If you guys want a better video forum, I really suggest reporting anyone you see accusing, attacking, baiting, flaming. You know the difference between right and wrong, so act on the wrong.
This thread was supposed to be informative and fun. The outcome truly does not matter as all the fun lies in actually doing the work and seeing and comparing results.
It's locked now, but I'd advise any of you to go visit PFI and read through the entire thread and see who is saying what. Thread was designed to rip me apart, but I don't care. Read it.

Now, can we get these benchmarks underway? Lets have some fun!

 

[Keysplayr]

If you run any demos I'd be happy to throw in a few maverick oc'd 5850 results, but my newer games library is pretty light-on at the moment (currently working through some Serious Sam HD, Tales of Monkey Island, FEAR and some casual NFS:HP2 and Colin McRae 2005

Sure Dug. Send me the captured fraps .csv files and I'll add them to the web results page. email is in the OP.

 

[evolucion8]

While I agree that HardOCP results are out of proportion (Heck, its easy to see that the CF scaling is pretty low compared to single GPU results), but it has nothing to do with this thread. I would love to see some benchmark results.

 

[Keysplayr]

While I agree that HardOCP results are out of proportion (Heck, its easy to see that the CF scaling is pretty low compared to single GPU results), but it has nothing to do with this thread. I would love to see some benchmark results.

How do you figure this?

 

[Madcatatlas]

How do you figure this?

You mean this has everything to do with that and thats why you wanted the benchmark scores from GTX460 SLI, Radeon HD5870 and Radeon HD5970?..baffling (turn on the sarcasm meter for this one gais)


Your previous testing of games when the GTX480 was released, did not help your credibility much Keys. You know this.
Now you make a topic where you include the TOP Radeons and pit them against the (by numbers) LOWEST Geforce.

It really shouldnt be a competition. And you know this.

The above are: observations. And although i couldnt care less, im trying to help you see why someone abit more into the red team, would have objections/opinions on your projects.

I would love to see some scores. especially with overclocked GTX460s, GTX470s and overclocked HD5870s , HD5850s and a HD5970

 

[Madcatatlas]

And for gods sake, dont get personal, im just voicing my observations. I have ordered myself a Gainward GTX460 GLH edition and im interested in the scores because of future updates!

Edit: congrats on 15k posts Keys!

 

[Keysplayr]

Ok I'll bite. What about my GTX480 testing?

 

[happy medium]

Ok I'll bite. What about my GTX480 testing?

I can't believe you took the bait. :\
Don't derail your own thread.

 

[Elfear]

Ok I'll bite. What about my GTX480 testing?

Keys I honestly think it's the source of the benchmarking. You may do the testing with the greatest degree of objectivity but because of your associations, your credibility will suffer. It's the reason why Nvnews and AMDZone are never high on the list of reliable review sites, even though their testing methodology might be just fine.

This thread is a little different in that you're getting independent members to run the benchmarks. I imagine though that it will still be seen as a less credible review just because the benchmarking came at the request of a focus group member whose results will be added to the mix.

Just my .02. Sounds kind of fun to me personally. If I had any of the latest games, I would love to contribute.

 

[Keysplayr]

Keys I honestly think it's the source of the benchmarking. You may do the testing with the greatest degree of objectivity but because of your associations, your credibility will suffer. It's the reason why Nvnews and AMDZone are never high on the list of reliable review sites, even though their testing methodology might be just fine.

This thread is a little different in that you're getting independent members to run the benchmarks. I imagine though that it will still be seen as a less credible review just because the benchmarking came at the request of a focus group member whose results will be added to the mix.

Just my .02. Sounds kind of fun to me personally. If I had any of the latest games, I would love to contribute.

I see what you're saying. But it's not my credibility that suffers. That only happens in the eyes of the beholders. I know my benches were always, and always will be done with the greatest care and accuracy. That credibility thing means nothing to me as I know I do the right thing. That's all that matters.

@ Happy. You're right. Bait spit out. LOL.
Onward!!!

 

[Annisman*]

Geez I just joined up here at Anandtech, and I am already amidst a brutal and bloody war of focus members/ fanboys / conspiracy theorists. Let's just bench some games and show the results!

Got a suggestion on how to bench the games: 1920x1080 res. 4XAA where applicable and 16AF, all very high settings....sounds simple to me ? Can we just bench all games at these settings ?

 

[Keysplayr]

Geez I just joined up here at Anandtech, and I am already amidst a brutal and bloody war of focus members/ fanboys / conspiracy theorists. Let's just bench some games and show the results!

Got a suggestion on how to bench the games: 1920x1080 res. 4XAA where applicable and 16AF, all very high settings....sounds simple to me ? Can we just bench all games at these settings ?

Sounds good to me. I'm just waiting on results to come on in.
Don't forget, collect two FRAPS .csv file. One for min max avg, and one for framerate. Go for it Annisman!!!

 

[evolucion8]

How do you figure this?

Simply compare recent reviews of the GTX 460 from other sites like Anandtech against the HardOCP results, you may surprise a little.

But definitively I would believe your benchmark results, in the end even if you lie (Something that I really doubt), you will only lie to yourself, not ourselves. But definitively the GTX 460 1GB SLI is a devasting force for 1920x1200 and below. Anything higher, the GTX 470 is the better buy.

 

[Keysplayr]

Simply compare recent reviews of the GTX 460 from other sites like Anandtech against the HardOCP results, you may surprise a little.

But definitively I would believe your benchmark results, in the end even if you lie (Something that I really doubt), you will only lie to yourself, not ourselves. But definitively the GTX 460 1GB SLI is a devasting force for 1920x1200 and below. Anything higher, the GTX 470 is the better buy.

Forget it dude. We've moved past you.