Quad-SLI and Quad-CrossFireX Interest?

Quad-GPU Poll

  • Yes I want to see this

    Votes: 135 71.1%

  • Not really interested

    Votes: 49 25.8%

  • Other

    Votes: 6 3.2%
  • Total voters
    190
i don't think i'd ever run quad sli/xfire. but i'd definitely like to see what the scaling/performance is since i prefer the way you guys tests cards at eyefinity/nvsurround resolutions vs most of the other sites that have done reviews on quad sli/xfire so far at low eyefinity/nvsurround res, non eyefinity/nvsurround res.

teletran8 said:
Dude I don't want to extrapolate anything you try to say :p But here goes!!

You ever notice [H] reviews will do things like this for example from the latest Graphics Benchmarks of the ASUS 6950 on Lost Planet 2.

1920x1080p = Min = 19 MAX = 52 Avg = 37.5 @ 4X MSAA 16X AF
2560x1600p = Min = 19 MAX = 50 Avg = 38.0 @ 0X MSAA 16X AF

LOL SO ALL THEY DID was drop AA when gaming at a higher rez? And I learned you gain performance by dropping AA like FREAKING DUH DUH BROSKI.

On the next page all I see is F1 2010 @ 2560x1600p which I don't game at that res I game @ 1080p so it make me less interested in reading beyond this point, because I won't know when I run that game what I'm really likely to see at my rez. All I can extrapolate is slightly more than 40 fps, which is at a point where some games are playable and some are slightly not, especially in more intense spots of a game so will it be playable for me and thousands of others that game @ 1080p? Who the heck knows lol! Is it that hard to change the res in options and give us sum numbers? lolwutlazy! They did it on the previous page...guess not they must be to to tired they gave up doing 1080p on the first page lol.

Next page Civ 5 same thing 2560 res and they gettin mins of 25 fps but peaks of 100fps what does this tell me? half the time the game runs choppy and other half the time it runs smooth or something? If I run it at 1080p do I get 30 or 40 fps or what for Minimums? Let me know sometime Mr. Numbnuts? What can you extrapolate from that review how CIv 5 will run @ 1080p for me? :p

Next Benchmark BC2. That game will run fine on any current Hardware. So that's unnecessary to review unless your in the 2 percentile that have 2560x1600 lmao!

I get all my Graphics reviews from Tech Power Up, I love the way they do reviews. @ all resolutions, but [H] does do a nice job of recommended playable settings, but I think they should stick to 1080p. Civ 5 25 fps min and 100 fps max what does that really tell anyone that doesn't game on a 2560x1600 it tells us SQUAT! Like you tell me here everyday :D
Click to expand...


just because you use a 1080p monitor doesn't mean the rest of us do so why should we be limited to your 1080p display.. anyone with half a brain can figure out what graphic settings you can use based on what works at 2560x1600, its not rocket science. i'd rather know what the absolute highest the card can run so i can figure out what will work for me. instead of having to look at reviews with limited settings like 4x AA/16x AF at the resolution i use even though the card can obviously do 8xAA/16xAF or higher at that specific resolution.

as far as civ 5 goes the frame rates in themselves are useless due to it being a turn based game, but what matters is the quality settings that are playable. with BC2 the more important thing is the driver updates from nvidia and AMD and if the frame rates improve with those new drivers since typically both AMD and Nvidia release new drivers with different tweaks for that card and certain games when new card is released. i won't even bother explaining lost planet 2 since you obviously don't know the difference between 1920x1080 4x AA and 2560x1600 0xAA at the pixel count difference.
 
Last edited:
If you have nothing else to do, sure, but I would rather see more reviews on cards that I might actually purchase.....
 
Not interested, it's like beating a dead horse already, the GTX 590 sucks and it will get it's ass handed to it in quad vs quad setup. If you can put this together fairly easily then go ahead for the hell of it, but everyone who knows their ass from their elbow knows what the results will be before you even go through the days of work to get it done.
 
Count me also as not interested. I think testing of the 6990 vs GTX 590 alone is enough.

Most people on these forums run a single GPU anyway, so why would they care about testing a $1400 video card combo? You'd need a $5000-6000 (total) system with 3x30" monitors to even consider that investment in the first place. I know a lot of you consider yourselves [H]ard, but VERY FEW are anywhere near that hardcore.
 
I'm interested. Not so much because of the results themselves, but because of the scaling the cards will do. The Anandtech article done properly.

With the AMD 6xxx series vastly improved cf scaling, this opens for new combinations of cards, to give better performance with lower cost. Trifire 5770 anyone?

Go for it if you have the time.

Thx for a fabulous site :)
 
Don't forget to put crysis 1 in the benchmarks people want to see the results. Forget about crysis 2 lol. Also how viable are external video cards? will thunderbolt "light peak" make a difference yet?
 
Interested? Hell yeah!

Relevant to me...heh, not likely! Although I may, just maybe, throw in a 6990 to go with my 6970 later this year and tri-fire it up.
 
NorthernEurope said:
I'm interested. Not so much because of the results themselves, but because of the scaling the cards will do. The Anandtech article done properly.

With the AMD 6xxx series vastly improved cf scaling, this opens for new combinations of cards, to give better performance with lower cost. Trifire 5770 anyone?

Go for it if you have the time.

Thx for a fabulous site :)
Click to expand...


the 5700 and lower series scaled perfectly, it was the 5800 series that had all the scaling issues. which is why its such a big deal that the 6900 series scales so well.
 
teletran8 said:
DO EEET IN 640X480P!!!

SVGA res lolwut?!

[H] writers HEY! take a look at Steam Hardware report sometime. http://store.steampowered.com/hwsurvey SEE WHERE IT SAYS PRIMARY DISPLAY RESOLUTION?!?! 1080P IS THE PRIMARY DISPLAY RESOLUTION, KNOCK IT OFF WITH ALL THAT OTHER RES NONSENSE PLEASE!!

I agree!! But do it @ 1080p, and a Triple Monitor 1080p resolution. Most people do not have 2560x1600 and they don't have 640x480 or 800x640 or 1280x1024 for that matter anymore IT'S 2011!!!!! (PLUS it doesn't help inform the mainstream who do have 1080p when you guys do these ABSURD Graphics and CPU benchmarks at these resolutions almost no one games in!). Also do one @ 1080p with no Anti Aliasing and one with Max Aliasing. Then you might have increased web traffic here lol! :p Or just start using Win 95 as your O.S. of choice because it shows performance increase or sumthing, DOH!. :D
Click to expand...
A quick comment to the mods about this issue. I think it will help people relate to the 1080p + rez if you state how many times larger the rez is to 1080p.How many 1080p screens will the rez fill...? do it like that. :)
 
Last edited:
Am I the only person who feels old and cantankerous when people say 1080p instead of 1920x1080? Just seems like another console encroachment dumbing down our communication.
 
We are doing it folks.... and we are using even newer drivers and profiles ;)

2x 6990 and 2x 590 :)

Stay tuned....
 
Please include 2560x1600 if you could. Just to see the raw power without any Vram limitations.
 
I think after the Tri-GPU review put out today that we can all come to a pretty good conclusion of what 4 Gpu's would look like, nevertheless i am interested to see the results in a professional controlled environment.

Anything you guys put out is awesome and informative, your one of if not the only site that does multi monitor benches.
 
obliv said:
I think after the Tri-GPU review put out today that we can all come to a pretty good conclusion of what 4 Gpu's would look like, nevertheless i am interested to see the results in a professional controlled environment.

Anything you guys put out is awesome and informative, your one of if not the only site that does multi monitor benches.
Click to expand...

The article put out today about 6990+6970 vs 580 Tri-SLI makes it easy to know the results of the upcoming 2X6990 vs 2X590.. probably. ;)

And yes for real tests made by professionnals in a controlled environment. Tired of reading ''tests'' made by anonymous Joe-6-pack living in a basement with grandpa, and posted everywhere on the internet like they are definitve results and answers to everything.

HardOCP is always the first place I come to read professionnal and well thought reviews. They do a great job.
 
[H] review please! also let us know the increase to your power bill with this one :)
 
I'm interested in this just to showcase the rumored improvements in performance with the latest (11.4?) ATI drivers. I would love to see at least a note comparing the single 6990 11.3 vs 6990 11.4 before the actual test with 2x 6990 and 2x 590 is done!

AMD/ATIs driver improvements have been stellar and it would be cool to see something about it. Tomshardware did two articles showcasing the importance of having the latest drivers here
Not site plugging but I think taking this into account is quite important when comparing cards that may have the best ability to benefit from these driver updates.

:)
 
I have a new CAP right now that is not released yet, should be next week, that we are using. You just don't get any more updated on drivers/profiles than us, it is a pet peeve of mine to make sure we test with the latest softwares.
 
Vega said:
The article would be the same old same old. Anyone who would be using 4x GPU in any combination would not be playing on anything less than a single 2560x1600 monitor. The POS 590s will run out of VRAM in any surround tests, get crippled and AMD crowned winner.

A much more interesting article would be to test 3GB 580s as they are now widely available. Removing any VRAM limit bottlenecks is the only way to truly test who has the best architecture and drivers/scaling.
Click to expand...

Agreed. Any test involving Surround and <1.5GB of VRAM will have a similar outcome to the one observed today. You guys at [H] would enlighten us a lot more by comparing something with ample VRAM, like 3GB 580s. I voted no in the poll, because frankly, spending the time to benchmark either configuration is a waste of time, but if you must bench an obscenely fast setup, make it one that won't run out memory.
 
XacTactX said:
Agreed. Any test involving Surround and <1.5GB of VRAM will have a similar outcome to the one observed today. You guys at [H] would enlighten us a lot more by comparing something with ample VRAM, like 3GB 580s. I voted no in the poll, because frankly, spending the time to benchmark either configuration is a waste of time, but if you must bench an obscenely fast setup, make it one that won't run out memory.
Click to expand...

Sigh. You and Vega should read the reviews again. :rolleyes:

I think it's clear, and well explained, that HardOCP are doing everything to keep the lack of VRAM on the 580 out of the way when testing.

Brent himself: '' Note that the settings I used were not memory capacity limited for the ap2ap tests, just saying, in these tests, memory was not holding back the performance''

Just read the reviews fully, it's well explained in there.
 
But this is better, you get your 'jollies', but don't have to spend money, unless you just love 'throwing dollars'. :D

Mr. Wolf said:
Me. I don't go to the Cheetah (in Atlanta) because no matter how much I may wish it or how money I slide between the garters and shapely thighs of those incredibly hot dancers, none of them will have sex with me. I consider these monster graphics card showdown threads to be the [H] version of the Cheetah. Just sayin'...
Click to expand...
 
Last edited:
This would be fantastic, but only if you stop using that 3.6Ghz 920!
 
Just for the fun of it. The highest 3D Mark 11 GPU score I have seen on the internet for Nvidia 2 X 590 is 19000. So we can already extrapolate the results from here, since my GPU score is 22452. :)

So AMD are also scaling well with higher CPU speed.. not just Nvidia. :)

quad53.jpg
 
I'd like to see this, but I think it would really be great if you tried to show how CPU dependent your results were. When guru3d.com did their equivalent review, they were often seeing little difference between the 6990 single card results and the 2x6990 crossfired results due to what they believed was CPU bottlenecking.

So, if you could perhaps test somewhere for a few different CPU scenarios (eg. a Phenom II [email protected] compared to an i5 [email protected] compared with an i5 SB chip), THAT would really be outstanding!

It would also be uber cool if you showed results for runs using 1x6990, 1x6990+1x6970, and 2x6990, just to let us see where the best bang for the buck lies. I know I'm asking for quite a bit, but you'd own the crossfire scaling gold medal for review sites on the internet if you did. :)
 
Hate to say it, but I expect the [H] to go into this review.


dooooo et!
 
This would probably be one of the most epic reviews ever, hopefully there could be a video review of it as well, that'd be even cooler, but I'm doubting that would happen, that'd be quite a production.
 
anubis44 said:
+1 to this!
Click to expand...

Funny guy. :) Bah. I'm missing those good old days. But we don't argue anymore, since the last 2 HardOCP reviews did the ''arguing'' for me. :D

And while I was saying that 6990+6970 was better then 580 SLI, they were calling me names, etc. But I was all wrong, since now, we all know, with the latest HardOCP review, that 6990+6970 can even take 580 Tri-SLI without any problems.

So I was totally wrong when saying ''6990+6970 is better then 580 SLI for 1000$, 2 cards against 2 cards''. I should have said ''6990+6970 is better then 580 Tri-SLI, for 500$ less and 200W less''. Know I know better. ;)

No need to argue anymore. Sometimes, the silence is deafening.
 
Well, it'd be no surprise if AMD won this one by a large margin. Only really interesting part is how much the fourth GPU will scale, for either side. Need to be careful with the 590's vRAM limit though, not much useful results if that's keeping it down.
 
I think you may be asking the wrong question. "Do you want to see X or not?" naturally tends to receive more yes than no answers.

Maybe you should instead ask:
Which of the following would you rather like to see first?
  1. 590 Quad-SLI vs. 6990 Quad-Fire
  2. 580 Quad-SLI vs. 6970 Quad-Fire
  3. 570 Tri-SLI vs. 6950 Tri-Fire
  4. 580 3GB SLI vs. 6970 Tri-Fire << my vote
  5. 580 3GB Tri-SLI vs. 6990 Quad-Fire
  6. etc.
(The Nvidia setups are more expensive, but somehow I suspect that they won't be much faster, if at all :D)

Too bad that the rig which Kyle is going to send you does not have 4 PCIe x16 slots, so 2. won't be possible
 
NukeDukem said:
I just want to see Vega and LEVESQUE argue some more :D
Click to expand...

It is funny that you mention this. Levesque happens to have bought a Quad AMD setup and three of the same monitors I have. I have challenged him to a benchmark duel. Who would like to see that!
 
XacTactX said:
Agreed. Any test involving Surround and <1.5GB of VRAM will have a similar outcome to the one observed today. You guys at [H] would enlighten us a lot more by comparing something with ample VRAM, like 3GB 580s. I voted no in the poll, because frankly, spending the time to benchmark either configuration is a waste of time, but if you must bench an obscenely fast setup, make it one that won't run out memory.
Click to expand...

Agree.
 
Vega said:
It is funny that you mention this. Levesque happens to have bought a Quad AMD setup and three of the same monitors I have. I have challenged him to a benchmark duel. Who would like to see that!
Click to expand...

A challenge. A duel! LOL! I don't know how old you are, but we are not figthing for the cutest little girl in our 6th grade class. But maybe you are? :D

But if everything is working, it will be my pleasure to make you feel better about your expenses, in comparing your 11K$ system with a system costing 1/2. So you will be able to feel better about all the money you spent on your computer.

If it makes you feel better, I will do it, just for the fun of it. Me, I pay for this with a small portion of the intererests I make on well placed money, so I don't need to justify my expenses.

But AMD are telling me your Crossfire bridge story without enough bandwidth is false. Maybe they are wrong, and you are right. I don't know, honestly. But I will know for sure soon.

BTW, AMD did ask me alot of questions about you and your claims, with links to your posts, and I think they will try to find you in real life with your ISP. You should be careful. Just saying. ;) Since you did post that on 50 different forums... Never know what they will do about all this....

But if you were right, I will tell everyone, in all honesty. And will be the first to say I was wrong.

But a duel! 5K$ set-up against a 11K$ set-up! You're funny. I will try yo find some time in my busy schedule to help you feel better. No problems. It's like Once upon a time in the West!
 
You must log in or register to reply here.
Back
Top