RD580

[Janooo]

According to the famous Inq:

"We were shocked to learn that RD580 will end up faster significantly than RD480 as once we moved from Nforce 4 SLI to Nforce 4 SLI 32 PCIe lanes we only got some marginal performance increase. I guess ATI did a great job with RD580 and it works much better than RD480 and partners like Asus decided to do a high end board based on the design. ATI's Northbridge features 36 lanes while Nvidia has 20 lanes in Northbridge and twenty in Southbridge. "

If I see it right it's like 10%-20% increase. Could it be possible?

Who is going to re-test all the PCIe cards for the new numbers?

 

[mwmorph]

anadtech has a review up already. best ocing board i have ever seen.

 

[Ackmed]

Hard, got a pretty good boost; http://enthusiast.hardocp.com/article.html?art=OTY5LDEsLGhlbnRodXNpYXN0

 

[nitromullet]

Originally posted by: Ackmed
Hard, got a pretty good boost; http://enthusiast.hardocp.com/article.html?art=OTY5LDEsLGhlbnRodXNpYXN0

Over what though..? The Inq so far is the only website to compare the A8R32-MVP with the A8R-MVP. Comparisons to other CrossFire motherboards with two cards is what I would really like to see more of. I'm sure there will be a roundup by someone with all the new RD580 mobo in spring.

Right now I'm liking my new A8R-MVP, although I've only had it for a short while so I won't put that into stone yet until I've run it through it's paces.

 

[Janooo]

I don't think it's so simple to get that kind of increase. The Inq must have made a mistake.

 

[KeithTalent]

Firing Squad

Quote:

We use the word ?potential? because although on paper the CrossFire Xpress 3200?s dual x16 graphics slots theoretically deliver twice the peak bandwidth to two Radeon X1900 graphics cards running in CrossFire mode, our performance results with actual applications showed little performance improvement from the added bandwidth for the most part. The only applications that really showed an advantage were 3DMark 06 and surprisingly, Call of Duty 2 rather than F.E.A.R. We saw a performance improvement of up to 9% in Call of Duty 2, while our 3DMark 06 overall score improved by just under 2%. Keep in mind that this higher score was based largely on the performance of one game test ? Canyon Flight ? one of two tests to utilize HDR lighting extensively.

Seems to contradict the INQ, shocker

 

[Matt2]

I'll trust firing squad over the Inq any day.

Sometimes it just seems the Inq posts crap to get people to read their articles.

 

[KeithTalent]

Originally posted by: Matt2

Sometimes it just seems the Inq posts crap to get people to read their articles.

Agreed, it seems to be working though.

 

[Hail The Brain Slug]

Notice, Hard has the FX-60 in the A8R32-MVP, and a 4800+ and a 4000+ in the other board. I would expect an increase from both a 4000+ and a 4800+ to an FX-60.

 

[Janooo]

Originally posted by: KeithTalent
Firing Squad

Quote:

We use the word ?potential? because although on paper the CrossFire Xpress 3200?s dual x16 graphics slots theoretically deliver twice the peak bandwidth to two Radeon X1900 graphics cards running in CrossFire mode, our performance results with actual applications showed little performance improvement from the added bandwidth for the most part. The only applications that really showed an advantage were 3DMark 06 and surprisingly, Call of Duty 2 rather than F.E.A.R. We saw a performance improvement of up to 9% in Call of Duty 2, while our 3DMark 06 overall score improved by just under 2%. Keep in mind that this higher score was based largely on the performance of one game test ? Canyon Flight ? one of two tests to utilize HDR lighting extensively.

Seems to contradict the INQ, shocker

I see only Crossfire comparison. There is no single card chart.

 

[KeithTalent]

Originally posted by: Janooo

Originally posted by: KeithTalent
Firing Squad

Quote:

We use the word ?potential? because although on paper the CrossFire Xpress 3200?s dual x16 graphics slots theoretically deliver twice the peak bandwidth to two Radeon X1900 graphics cards running in CrossFire mode, our performance results with actual applications showed little performance improvement from the added bandwidth for the most part. The only applications that really showed an advantage were 3DMark 06 and surprisingly, Call of Duty 2 rather than F.E.A.R. We saw a performance improvement of up to 9% in Call of Duty 2, while our 3DMark 06 overall score improved by just under 2%. Keep in mind that this higher score was based largely on the performance of one game test ? Canyon Flight ? one of two tests to utilize HDR lighting extensively.

Seems to contradict the INQ, shocker

I see only Crossfire comparison. There is no single card chart.

That is true, so it's not exactly oranges to oranges (maybe kumquats to oranges) , however, seeing the Inquirer's increase of 10-20% on a single card, one would expect something similar (heck, even half that) from other reviewers, I was just pointing out that is not the case right now. Maybe someone else will come out with a review to validate The Inquirer's numbers, but I will not hold my breath.

 

[nitromullet]

Originally posted by: Janooo

Originally posted by: KeithTalent
Firing Squad

Quote:

We use the word ?potential? because although on paper the CrossFire Xpress 3200?s dual x16 graphics slots theoretically deliver twice the peak bandwidth to two Radeon X1900 graphics cards running in CrossFire mode, our performance results with actual applications showed little performance improvement from the added bandwidth for the most part. The only applications that really showed an advantage were 3DMark 06 and surprisingly, Call of Duty 2 rather than F.E.A.R. We saw a performance improvement of up to 9% in Call of Duty 2, while our 3DMark 06 overall score improved by just under 2%. Keep in mind that this higher score was based largely on the performance of one game test ? Canyon Flight ? one of two tests to utilize HDR lighting extensively.

Seems to contradict the INQ, shocker

I see only Crossfire comparison. There is no single card chart.

First of all. no surprise that the Inq's results don't mesh with a reputable site's...

Second, why would you buy an RD480 or RD580 board to run a single card at this point? There are a lot of much more proven and feature rich motherboards out there for the money for single card rigs.

 

[nib95]

Originally posted by: nitromullet

Originally posted by: Janooo

Originally posted by: KeithTalent
Firing Squad

Quote:

We use the word ?potential? because although on paper the CrossFire Xpress 3200?s dual x16 graphics slots theoretically deliver twice the peak bandwidth to two Radeon X1900 graphics cards running in CrossFire mode, our performance results with actual applications showed little performance improvement from the added bandwidth for the most part. The only applications that really showed an advantage were 3DMark 06 and surprisingly, Call of Duty 2 rather than F.E.A.R. We saw a performance improvement of up to 9% in Call of Duty 2, while our 3DMark 06 overall score improved by just under 2%. Keep in mind that this higher score was based largely on the performance of one game test ? Canyon Flight ? one of two tests to utilize HDR lighting extensively.

Seems to contradict the INQ, shocker

I see only Crossfire comparison. There is no single card chart.

First of all. no surprise that the Inq's results don't mesh with a reputable site's...

Second, why would you buy an RD480 or RD580 board to run a single card at this point? There are a lot of much more proven and feature rich motherboards out there for the money for single card rigs.

hense the reason I bought the Asus A8N32 SLI Deluxe instead. I know it's still SLI, I honestly wanted to buy the DFI Lanparty Ultra-D, but it was the Asus boards rich feature set and external SATA port that stole it for me.

 

[Elfear]

Originally posted by: nitromullet


Second, why would you buy an RD480 or RD580 board to run a single card at this point? There are a lot of much more proven and feature rich motherboards out there for the money for single card rigs.

For the overclocking features. Hopefully the RD580 chipset will allow for a higher fsb than current NF4 boards. The Asus isn't looking too promising in that department but my fingers are crossed that the DFI will run like a chimp on crack.

 

[mwmorph]

Originally posted by: Elfear

Originally posted by: nitromullet


Second, why would you buy an RD480 or RD580 board to run a single card at this point? There are a lot of much more proven and feature rich motherboards out there for the money for single card rigs.

For the overclocking features. Hopefully the RD580 chipset will allow for a higher fsb than current NF4 boards. The Asus isn't looking too promising in that department but my fingers are crossed that the DFI will run like a chimp on crack.

? It's the best ocing board anandtech has ever tested. 322mhz at 1t is the highest at 1t. The Asus Sli deluxe does better but needs 2t, albeit the 2t mode is as fast as most other boards' 1t mode. As for stock multiplier ocing, it got to 246 at 1t, better than the other 246 Asus NF4 sli Deluxe which did 246 at 2t. Dfi is not king of the ocing pile and it's boards have compatibility problems. Why not the asus for overclocking?

 

[nitromullet]

Originally posted by: Elfear

Originally posted by: nitromullet


Second, why would you buy an RD480 or RD580 board to run a single card at this point? There are a lot of much more proven and feature rich motherboards out there for the money for single card rigs.

For the overclocking features. Hopefully the RD580 chipset will allow for a higher fsb than current NF4 boards. The Asus isn't looking too promising in that department but my fingers are crossed that the DFI will run like a chimp on crack.

But you should still add a second video card...

 

[Elfear]

Originally posted by: mwmorph

? It's the best ocing board anandtech has ever tested. 322mhz at 1t is the highest at 1t. The Asus Sli deluxe does better but needs 2t, albeit the 2t mode is as fast as most other boards' 1t mode. As for stock multiplier ocing, it got to 246 at 1t, better than the other 246 Asus NF4 sli Deluxe which did 246 at 2t. Dfi is not king of the ocing pile and it's boards have compatibility problems. Why not the asus for overclocking?

Well 322MHz is only 4MHz higher than the boards that have been out for a year (as tested by Anandtech). That isn't too phenomenal in my book. Personally, I'm looking for a board that will do ~350MHz so I can let my cpu stretch it's legs a bit. I'm stuck with a x9 multi so I need a board that will do crazy fsb numbers. That's why i'm not interested in the Asus.

As far as DFI's being king of the hill, have you looked at any of the overclocking records lately? DFI has a huge presence there and it's not by accident. I went through 5 different boards trying to find a decent overclocker before I settled on my current DFI. There are a few other boards out that are comparable, but they are mostly recent editions to the motherboard arena. DFI has been pretty much top dog for along time since the Ultra-D (and variants) came out. The Expert exceeded the Ultra-D by bit and is at least in the top three of overclocking boards, if not #1.

Originally posted by: nitromullet

But you should still add a second video card...

Lol. Very true. But should and could are two different things.

 

[deepinya]

Originally posted by: nitromullet

Originally posted by: Elfear

Originally posted by: nitromullet


Second, why would you buy an RD480 or RD580 board to run a single card at this point? There are a lot of much more proven and feature rich motherboards out there for the money for single card rigs.

For the overclocking features. Hopefully the RD580 chipset will allow for a higher fsb than current NF4 boards. The Asus isn't looking too promising in that department but my fingers are crossed that the DFI will run like a chimp on crack.

But you should still add a second video card...

Does your board limit the refresh rate as did the rd480 board?

 

[Matt2]

Was it the actual RD480 that limited the refresh rate? I thought that was only a limitation in X800 crossfire?

 

[Elfear]

Originally posted by: Matt2
Was it the actual RD480 that limited the refresh rate? I thought that was only a limitation in X800 crossfire?

That was my understanding too. Seems like I've seen some reviews using Crossfire and the RD480 together at resolutions above 1600x1200.

 

[Zebo]

Myth.. AMD makes the only part of the chipset that matters, AMD Hammer IMC. Therefore it's no accident ALL hammer boards perform more less identically. SiS, Uli, NF3, via, NF4, 480, 580 whatever does'nt matter. Last year and a half I've never seen sets of boards with benchmarks so close.

Outliers, well within margins of errors.. are accounted for by, lack of repeating and averaging of benchmarks by time pressed reviewers, bios difference, trace differences .. basically individual board quality and EE smartness.

 

[deepinya]

Originally posted by: Elfear

Originally posted by: Matt2
Was it the actual RD480 that limited the refresh rate? I thought that was only a limitation in X800 crossfire?

That was my understanding too. Seems like I've seen some reviews using Crossfire and the RD480 together at resolutions above 1600x1200.

Im referring to refresh rate of the monitor, not resolutions.

x1800xt on the rd480 had the same issue. Is it the rd480 chipset or all crossfire cfgs?

 

[Elfear]

Originally posted by: deepinya

Im referring to refresh rate of the monitor, not resolutions.

x1800xt on the rd480 had the same issue. Is it the rd480 chipset or all crossfire cfgs?

I hadn't heard that. The only problem I remember involving refresh rates was with the X850 series cards in Crossfire being limited to 1600x1200@60Hz.

 

[fierydemise]

The refresh rate limitations were a problem on the X8x00 series cards, I also hadn't heard of the issue with all RD480 motherboards, I know AT(and others) have tested X1x00 crossfire on rd480 boards without problems so I think you're slightly mistaken.

 

[nitromullet]

Originally posted by: deepinya

Originally posted by: Elfear

Originally posted by: Matt2
Was it the actual RD480 that limited the refresh rate? I thought that was only a limitation in X800 crossfire?

That was my understanding too. Seems like I've seen some reviews using Crossfire and the RD480 together at resolutions above 1600x1200.

Im referring to refresh rate of the monitor, not resolutions.

x1800xt on the rd480 had the same issue. Is it the rd480 chipset or all crossfire cfgs?

There was a refresh rate limitation issue at resolutions 1600x1200 and higher with the X850 series CrossFire composting chip (which is on the video card itself), not the RD480 chipset. My rig doesn't have this issue, and the motherboard I have is based on RD480 and not RD580.