AMD Radeon R9 290X CrossFire Performance

[Vitalius]

It is NOT anywhere close to be fixed, they just reduced it a bit. To tell the truth, it will never be "fixed", because we don't live in the perfect world. HW fix that NVidia deployed would be nice and much better than driver, but as @Vitalius said not even that will remove it completely.

You are right it will never be perfect, but for all intents and purposes it is "fixed" now compared to what it was before. At the point it is now most people shouldn't notice the stuttering. People sensative to it still will of course, but some of those people can even notice it even on SLI so it's easier to measure to the point where it is fine for most people.

This basically^

It's not fixed the way SLI has it, but it wasn't acceptable before at all. Now it is plenty acceptable to play on, which in my mind is "fixed". At least up to 1080p (or was it 1440p?).

[Derangel]

This basically^

It's not fixed the way SLI has it, but it wasn't acceptable before at all. Now it is plenty acceptable to play on, which in my mind is "fixed". At least up to 1080p (or was it 1440p?).

 

2560x1600. It doesn't work for DX9 or Eyefinity.

[Altecice]

Merged Topic OP - 

 

Some charts have been leaked showing the performance gains from Single > CrossfireX configuration.

achieving up to 198%/200% gain on some games... have to say thats impressive! 

 

 

 

 

 

 

The full Article (and more charts) can be found HERE!

[heineygoat]

This graph is misleading because it doesn't start at 0.

At first glance, it appears to be about a 120% performance increase from Xfire.

[Altecice]

This graph is deceiving and somewhat bias because it doesn't start at 0.

At first glance, it appears to be about a 120% performance increase from Xfire.

I believe it just to fit the graph on most monitors, as you can see the line at 1.0 clearly shows where the graph really starts. If you read further into the Article you will see them compair them on pure FPS numbers over a GTX780.

[Imaginaerum]

I believe it just to fit the graph on most monitors, as you can see the line at 1.0 clearly shows where the graph really starts. If you read further into the Article you will see them compair them on pure FPS numbers over a GTX780.

To fit on monitors? That's not a reason, just resize it like with anything else. It's done to make it look inflated, but it does look promising. I wonder how the microstutter is.

[heineygoat]

I believe it just to fit the graph on most monitors, as you can see the line at 1.0 clearly shows where the graph really starts. If you read further into the Article you will see them compair them on pure FPS numbers over a GTX780.

The image is 850 pixels wide, I can't see how fitting it on any monitor would be a concern, or how that would be a good reason to generate a deceiving graph. Any graph that doesn't start at the origin is misleading. A graphs main purpose is to communicate data quickly by visual representation and at first glance it appears to be more than double performance from Xfire.

 

Obviously you didn't make the graph and it's still good information so thanks for sharing OP 

[helping]

Guys, keep in mind the chart starts a 0.4 and ends at 2.0

 

 

how about we talk about the results instead of the chart, near linear scaling in FC3 is pretty cool. 

[LinusTech]

They say this every generation.

 

So does NVIDIA.

 

It's never been true, so colour me surprised if it is this time.

[OlekKing]

They say this every generation.

 

So does NVIDIA.

 

It's never been true, so colour me surprised if it is this time.

Oh please our beloved Lord, tell us when the hell we'll see a review from you. I'm dying from anxiety waiting for these pieces of silicone to come out...oh please may you have mercy...

[Tall Gora]

They say this every generation.

 

So does NVIDIA.

 

It's never been true, so colour me surprised if it is this time.

i agree Linus. i have yet to see 2x scaling, that will be the day...

[LinusTech]

i agree Linus. i have yet to see 2x scaling, that will be the day...

 

I've seen it before. Witcher 2 scales GREAT on NVIDIA for example...

 

But these graphs that both AMD and NVIDIA show off are always at some kind of weird cherry-picked setting that usually doesn't end up representing what a normal gamer will encounter.

 

Perhaps GeForce Experience and to a lesser extent Raptr will help with this, but I'm going to take a "wait & see" approach to that too.

[ViruZ_HUN]

Some charts have been leaked showing the performance gains from Single > CrossfireX configuration.

achieving up to 198%/200% gain on some games... have to say thats impressive! 

 

 

 

 

 

 

The full Article (and more charts) can be found HERE!

 

if its true then r9 290x is impressing, it preforms around 10+ more in some games at 2160p

[EChondo]

I SWEAR, I've seen this vary same graph on these forums before.

 

OT: I hope this is how well it scales, but I am entirely doubtful. Could AMD have possibly fixed a specific issue in the past with Crossfire? I doubt it, but it's a possibility.

 

Just release the cards already AMD...

[Tall Gora]

I've seen it before. Witcher 2 scales GREAT on NVIDIA for example...

 

But these graphs that both AMD and NVIDIA show off are always at some kind of weird cherry-picked setting that usually doesn't end up representing what a normal gamer will encounter.

 

Perhaps GeForce Experience and to a lesser extent Raptr will help with this, but I'm going to take a "wait & see" approach to that too.

interesting, except the "wait and see" part is starting to become a little lengthy

[Altecice]

I've seen it before. Witcher 2 scales GREAT on NVIDIA for example...

 

But these graphs that both AMD and NVIDIA show off are always at some kind of weird cherry-picked setting that usually doesn't end up representing what a normal gamer will encounter.

 

Perhaps GeForce Experience and to a lesser extent Raptr will help with this, but I'm going to take a "wait & see" approach to that too.

 

AMD believe they have fixed the Microstutter because the R9-290X doesn't use a software method (via the ribbon cable) instead they now use the PCI-E slot themselves to CrossfireX which is now a Hardware solution and imo Hardware > Software. It clearly shows further into the article how they almost get a 200% scale. I agree however they cherrypick the games that are designed for AMD (Tomb Raider and Crysis3 come to mind) but it still looks like very good results. I know i am waiting for the 290x's to come out so i can fit two into my Eyefinity rig

 

Although i often wonder the difference between PCI 2.0 > PCI 3.0 with theese on crossfire as my setup is running the Crosshair V Formula-Z

 

Btw your streaming on Twitch at the Nvidia Conference (G-sync) was great.Keep eating that fruit haha.

[mgsstar]

TechPowerUp released a crossfire review: http://www.techpowerup.com/reviews/AMD/R9_290X_CrossFire/

[beebskadoo]

Over 80%? Awesome. Even in games that aren't Gaming Evolved titles? Extremely awesome. 

Well, looks like I may have made a good choice in aiming for Crossfire 290X's in the future for 4K. 

But yeah, agreed with lachyman. Frame timing is actually everything in regards to FPS. Otherwise it's just a number.

PCPER said the frametiming issues are less. But the issues are pretty bad in games like skyrim (and older titles) and 4k/surround gaming. 

[Vitalius]

PCPER said the frametiming issues are less. But the issues are pretty bad in games like skyrim (and older titles) and 4k/surround gaming. 

In regards to 290X crossfire? Or just AMD crossfire in general? Link please. And thanks.

[beebskadoo]

In regards to 290X crossfire? Or just AMD crossfire in general? Link please. And thanks.

I'm on mobile so it's kind of a pain. But it's on their site@ pcper.com

The 290x crossfire.

The frame timing issues are a lot better but still are not that great with surround and 4k gaming. And some older titles that are dx9

The implementation of the new drivers along with the new crossfire tech on the 290x helps a lot I guess hope this helps

Sent from my SGH-T999 using Tapatalk