30inch monitor.. SLI 680's or a 690?

[felang]

If you don't count microstutter and needing to use an fps limiter to fix it.

That's the only knock against dual AMD cards.

Yeah you´re probably right. I haven´t used a dual GPU setup in a few years mostly because IMO it´s not worth the hassle, but with a 30 inch monitor I really don´t see another option if you want to turn the eyecandy up anyway. But yeah, I do hear better things about SLI than crossfire.

 

[moonbogg]

I read a quick review on some prog named Radeon Pro and it reduces micro stutter on AMD cards apparently. Without it, the microstutter is supposed to be noticeably worse than SLI.

 

[Ferzerp]

Due to the crossfire issues. Dual 680.

People like to push AMD regardless, even when every reviewer out there states that for dual GPU, even when the AMD cards benchmark higher, the nvidia cards have a lower standard deviation in frame time so they end up being better.

For single card, it's mostly a wash with the AMD solution with a slight lead.

 

[cmdrdredd]

I read a quick review on some prog named Radeon Pro and it reduces micro stutter on AMD cards apparently. Without it, the microstutter is supposed to be noticeably worse than SLI.

It uses a frame rate limiter to do it.

 

[MrK6]

Dual GTX 680's is an absolute waste of money. GTX 670's +/- an overclock will offer the same gameplay experience for significantly less. If you do feel like burning money, at least a GTX 690 offers some tweaks to SLI (hardware frame metering, etc.) which will enhance the overall experience.

 

[AdamK47]

Due to the crossfire issues. Dual 680.

People like to push AMD regardless, even when every reviewer out there states that for dual GPU, even when the AMD cards benchmark higher, the nvidia cards have a lower standard deviation in frame time so they end up being better.

For single card, it's mostly a wash with the AMD solution with a slight lead.

How many CrossFire setups have you had first hand experience with?

 

[Ferzerp]

How many CrossFire setups have you had first hand experience with?

Are you implying that AT, [H], etc are just lying to us?

 

[AdamK47]

Are you implying that AT, [H], etc are just lying to us?

I take that as none.

 

[Ferzerp]

I take that as a yes.

 

[AdamK47]

Just deflecting the deflection.

Since you asked nicely, I have owned the following in multi-GPU configurations:
Two GTX 280s
Two GTX 285s
Three 5870s
Three GTX 480s
Three GTX 580s
and now three 7970s

I've seen micro stutter with both CrossFire and SLI. Some games will show it more on nVidia, some ATI. I find that they trade blows in this regard. One of the best ways to minimize it for any form of AFR is to maintain 60 FPS with vsync (or what your refresh rate is with vsync). It solves the problem 99% of the time.

 

[Ferzerp]

You are trying to imply that we should ignore various reviewers and listen to you instead. You're also trying to shift the burden of proof off yourself when your claims run contrary to what is widely published.

Furthermore, you're trying to refute what reviewers are saying by suggesting that for the reviewers to be correct, I must waste money on crossfire and verify it myself. Obviously, this is a ludicrous argument. What I own has absolutely no bearing on the accuracy of what reviewers may say. That's a pretty sad way to try to "win" a discussion.

 

[AdamK47]

No, I'm telling you from actual user experience by owning such hardware. I'm not trying to win anything.

Besides you made the statement, "the nvidia cards have a lower standard deviation in frame time so they end up being better". The wording alone is rather peculiar.

It would have been better if you really did have the sources linked in your original reply, but I guess that wouldn't really fit in with all of the other off the cuff responses people generally make on these forums.

 

[Ferzerp]

Sure you're trying to win.

You are trying to get us to believe you instead of the larger body of evidence that disagrees with you. You're asserting that your claims are correct, and that reviewers are inept, misleading, or otherwise reporting incorrect information.

 

[AdamK47]

I'm not sure you read what I posted.

You say that I refute whatever it is this larger body of evidence is. Where have I refuted anything? Is it this: "I've seen micro stutter with both CrossFire and SLI. Some games will show it more on nVidia, some ATI. I find that they trade blows in this regard."?

I apologize for my poor refuting.

 

[Grooveriding]

When I was playing Guild Wars 2, nvidia's SLI was not working for over a month until they fixed it, so I simply disabled it.

Now I've never been much for having problems with microstutter, been using dual cards regularly since 280SLI, then 5870CF, then 480SLI, then 480 Tri-SLI and now 680 SLI. I've seen microstutter on all of them but mostly only in certain titles and generally only if I am not using vsync. Otherwise I always say I don't even notice it.

Well the fact is you just get used to it and any multi-gpu setup exhibits it. When GW2 started to have working SLI and I turned it back on I immediately noticed a big difference and saw micro stuttering. This was with vsync on as well.

This AMD vs nvidia microstutter garbage is just more forum fodder for one fan or another to make unsustainable claims over. If you use multi gpu you have microstutter. I've never seen a more shocking level of microstutter using one brand or another and hardly ever notice it because I just get used to it.

If you're running SLI go ahead and turn it off and play with one card for a week or so, then turn it back on. It's night and day until you get used to it again. I'm sure the same is still true of Crossfire as well.

 

[cmdrdredd]

Just deflecting the deflection.

Since you asked nicely, I have owned the following in multi-GPU configurations:
Two GTX 280s
Two GTX 285s
Three 5870s
Three GTX 480s
Three GTX 580s
and now three 7970s

I've seen micro stutter with both CrossFire and SLI. Some games will show it more on nVidia, some ATI. I find that they trade blows in this regard. One of the best ways to minimize it for any form of AFR is to maintain 60 FPS with vsync (or what your refresh rate is with vsync). It solves the problem 99% of the time.

Kepler (GTX 670/680) has hardware frame metering to limit microstutter. The solutions you listed all lack this functionality. You do not need to use vsync or any framerate limiters to reduce microstutter with Kepler GPUs.

I'm not saying you won't notice it but if you take X-Fire 7970s vs SLI 680s side by side with no frame limiters the GTX 680 will show significantly less microstutter and you'd have to limit your FPS on the 7970 setup to achieve the same results.

 

[tweakboy]

Depends what drivers, newer drivers favor ATI.



Still giving awesome advice I see...

Hell ya! vsync on 1ms ,,,,, smooth as your butter in the morning. instant no ghosting,,,, a 30 mon is laggy with vsync ON.

with OFF screen tearing is nasty nast nasty!!!!!!!!!

:whiste:

 

[cmdrdredd]

Hell ya! vsync on 1ms ,,,,, smooth as your butter in the morning. instant no ghosting,,,, a 30 mon is laggy with vsync ON.

with OFF screen tearing is nasty nast nasty!!!!!!!!!

:whiste:

Do you even speak english?

 

[AdamK47]

Hell ya! vsync on 1ms ,,,,, smooth as your butter in the morning. instant no ghosting,,,, a 30 mon is laggy with vsync ON.

with OFF screen tearing is nasty nast nasty!!!!!!!!!

:whiste:

My setup with a 30" monitor and three 7970s in CrossFire is supposedly a laggy stuttering mess then. That's what I'm getting from this thread.

 

[cmdrdredd]

My setup with a 30" monitor and three 7970s in CrossFire is supposedly a laggy stuttering mess then. That's what I'm getting from this thread.

I said more microstutter, not a mess. Perhaps it would be best if you read what was said instead of interpreting it in your own way.

As for tweakboy...i don't think anyone takes him seriously.

 

[AdamK47]

I said more microstutter, not a mess. Perhaps it would be best if you read what was said instead of interpreting it in your own way.

As for tweakboy...i don't think anyone takes him seriously.

I didn't interpret your post in my own way since I wasn't referring to it.

 

[cmdrdredd]

I didn't interpret your post in my own way since I wasn't referring to it.

Seemd like it...my apologies.

 

[Phishy714]

RAWR

 

[Ferzerp]

He's angry with me because reviewers rate the current generation of SLI higher than Crossfire. Because of that, he's trying to ridicule me for pointing out their words.

It's not my fault that SLI works better than Crossfire. If you want to get annoyed at anyone, get annoyed at AMD.

 

[AdamK47]

Let me get something straight here. This is the quote by Ferzerp that threw up a red flag.

People like to push AMD regardless, even when every reviewer out there states that for dual GPU, even when the AMD cards benchmark higher, the nvidia cards have a lower standard deviation in frame time so they end up being better.

His post makes it seem as if microstutter is a black or white thing. Either you have it when using CrossFire, or don't without CrossFire. It's not that simple. Hence the reason why I asked him if he actually used CrossFire before. Believe it or not, but you can indeed have a game running at 60 FPS on ATI hardware and have zero microstutter. If the same game running on nVidia hardware also exibits zero microstutter then this "standard deviation in frame time" (his quote) becomes essentially the same. Yes, there are cases on ATI hardware where you will notice it in some games and some CrossFire configurations, but it's not always going to happen and it's not going to be as critical as it's made out to be.