AMD / ATI Radeon HD5800 -review thread

I'm wondering how you're avoiding screen burn in.

 

Hardly relevant on new and "fresh" 1080p Plasma's. I've got an old 40" Plasma (Samsung) and it's already wrecked after 30 minutes of Guitar Hero

 

I have a low hour Samsung plasma that wants to burn in over any type of gaming I do. I really can't see how anyone 'games regularly' on one.

 

My mate has 0 problems on his 50" Samsung Plasma and he has his ps3 connected 24/7 but he spent a great deal of time calibrating the colors. As I said, I can't play for 30 minutes without burn, he has zero trouble even if we spent all night on Motorstorm

 

I broke the screen in very gently during the first few weeks of ownership. I still make it a habit to switch between video sources whenever I take a break from gaming. TBH, I don't game on it all that often anymore, at least not PC gaming. Console games have enough cutscenes that I don't have to worry about burn-in. Mine's a Panasonic model from 2008, supposedly by then they had worked out the burn-in issues.

 

As was said earlier multiple times, the new plasma models circa 2008 do not have issues with screen burn in. They may have slight image-retention, but after watching a different source the image retention is gone.

After a long GeoW2 Horde session, 6-8 hours with AlZilla and company, the HUD was barely visible on a black screen. It could only be seen as a slightly darker black shade when inspected from 1 foot away. After watching a movie or two it's completely gone. The screen is perfectly uniform again with no image-retention to be seen. This is on a Panasonic V10 2009 model.

 

Alright that's it. I'm plugging in my Dreamcast to play some Soul Calibur.

 

Eyefinity performance at [H]

http://hardocp.com/article/2010/01/05/amds_ati_radeon_eyefinity_performance_review/1

 

Beyond3D: ATI Cypress Gaming Performance Analysis

Dragons, triangles, conflicts and......

So the second rasterizer can't help?

Do you know if the GDS is already in use for tessellation?

 

they are talking about the article here

http://forum.beyond3d.com/showthread.php?t=56297&page=3

 

And I hate them for this.

 

It's not limited by raster, most likely, but by setup in those points, since you have a whole bunch of small triangles.

The GDS holds tessellation factors.

 

k, thanks.

http://www.pcgameshardware.de/aid,7...d-103-Neue-Features-im-Test/Grafikkarte/Test/
"AMD gibt jedoch an, die 1.1-Version der Heaven-Demo genutzt zu haben," (AMD says, they used the 1.1 version of the Heaven Demo (for their tests)).
Do you know something about this mysterious Heaven 1.1 demo?

 

I've only heard about it recently, and in a different context (it's the variant that's likely to see a Linux release too). One can assume that they'll try to round off some of the current rough corners, but I don't know much about it.

 

I didn't understand from the results page where 15% came from. Is it a estimation of some sort?

 

Perfstudio outputs that data, so no, it's not an estimate...unless it's estimated there instead of being derived from counter monitoring

 

Did you do a clock scaling evaluation for the toughest frame of tessellation? To get an idea of engine versus memory bottlenecks in tessellation?

Jawed

 

It would be nice to see it compared with a 5770 as well.

 

Here is partial response for you Jawed:

1920x1200 DX11 high

1000/1300


1000/1200


1000/1050


1000/900


As you can see memory has almost no impact on tesselation performance.
19FPS with memory @900-1200 was jumping to 20FPS and back with 20FPS most of the time.
19FPS with memory @1300 was steady and I can explain it by E.C. kicking in because my card GDDR5 is a weak point.