Pretty much expected that... wonder how much it'll cost and how it does in Crossfire.

 

Quote:

Originally Posted by Chunky_Chimp Pretty much expected that... wonder how much it'll cost and how it does in Crossfire.

Check the source for pricing

 

Quote:

Originally Posted by XAslanX Check the source for pricing

I had a problem loading it at first, so I couldn't... but it loads now so I guess it was a temporary problem on my end. Price range is what I thought it'd be but I'd like to see what the actual sticker price is when Newegg and other places start stocking them.

 

wait 5830 is 175w and 5850 is 150w?

 

Quote:

Originally Posted by gerikoh wait 5830 is 175w and 5850 is 150w?

+1

 

They failed to mention the memory bus type, assuming its a 256 bit.

Yeah, I find it strange the 5830 has a higher TDP than the 5850 or either its a typo.

 

Wonder how this things going to do. I want to see it with just about -10% off a 5850.

 

i really hope that it's just a type

else, i might as well just add some cash and grab a 5850 since it's just priced at $260 here.

or maybe i should just wait for fermi for more price drops

 

This thing looks kind of weak. I don't know, I think I'll just grab a 5870 in a month.

EDIT:

Could make a sick crossfire card though.

 

5830 is a higher clocked part, tdp raises very fast with high clocks .Also keep in mind these are cut up dies , ATI wouldn't cut up good dies , these are likely lesser bins of chips,leakier.Half the ROP count is disappointing ,but i guess if it wasn't ,the distance between it and 5850 would have been to small .

 

Hmm, it looks like it might actually perform within 5% of the 5850....

My 5870 scaled pretty linearly with clockspeed and with an 8% bump in clock speed offsetting a 20% loss in shaders (which scale much less linearly)... this will be much closer than it looks.

EDIT: didn't notice the ROPs... oh yeah, that'll kill the high settings performance.

 

Quote:

Originally Posted by gerikoh wait 5830 is 175w and 5850 is 150w?

Higher clock speed takes more power.

 

Some one like to inform a tech noob what the RoP's are?

 

Quote:

Originally Posted by Acoma_Andy Higher clock speed takes more power.

At a factor of ~1w per 1MHz, and with half the ROP's and one fewer shader cluster?

I was considering picking up this card, but with half the ROP's I'm not nearly as sure anymore. I'll wait for benchmarks of course, but as it stands I'm not quite as excited about this card as I was before I found that little tidbit out.

 

Quote:

Originally Posted by Chimeracaust Some one like to inform a tech noob what the RoP's are?

http://en.wikipedia.org/wiki/Render_Output_Processor

A decent enough explanation I suppose.

 

Quote:

Originally Posted by Chimeracaust Some one like to inform a tech noob what the RoP's are?

Processed pixel fragments are stored in card memory ready to be resolved into completed screen pixels, for output onto your display. This task is handled by a GPU unit called the ROP. A modern GPU implements a number of ROPs, based on how likely the GPU is to be bottlenecked by pixel output, to perform the final tasks of rendering. As well as simply resolving and drawing pixels on your screen, the ROP hardware also performs a number of optimisations to save memory bandwith when reading and writing pixels to and from a framebuffer, such as colour compression (even saving 1 byte of colour data per pixel is a heady saving in bandwidth terms).

The ROP units also deal with depth compression and compare, the compare - where you test pixels against each other to see which is on top of the other - being the main facilitator for multisample antialiasing. Multisample antialiasing uses depth information (Z) to alter the colour of pixels so that geometry rasterised earlier in the render process is antialiased and looks better.

http://www.hexus.net/content/item.php?item=17992

Basically , the draw back of having less ROP's is you might be bottle necked at higher resolution with high AA levels .

 

Just give me damn 6970 already.

 

Quote:

Originally Posted by RayvinAzn http://en.wikipedia.org/wiki/Render_Output_Processor A decent enough explanation I suppose.

Quote:

Originally Posted by ~CS~ Processed pixel fragments are stored in card memory ready to be resolved into completed screen pixels, for output onto your display. This task is handled by a GPU unit called the ROP. A modern GPU implements a number of ROPs, based on how likely the GPU is to be bottlenecked by pixel output, to perform the final tasks of rendering. As well as simply resolving and drawing pixels on your screen, the ROP hardware also performs a number of optimisations to save memory bandwith when reading and writing pixels to and from a framebuffer, such as colour compression (even saving 1 byte of colour data per pixel is a heady saving in bandwidth terms). The ROP units also deal with depth compression and compare, the compare - where you test pixels against each other to see which is on top of the other - being the main facilitator for multisample antialiasing. Multisample antialiasing uses depth information (Z) to alter the colour of pixels so that geometry rasterised earlier in the render process is antialiased and looks better. http://www.hexus.net/content/item.php?item=17992 Basically , the draw back of having less ROP's is you might be bottle necked at higher resolution with high AA levels .

I get it know! And thanks, +rep both of you.

 

think i'll add one to my 5850 one day.

 

So it should be about 40-45% faster then a 4890

More texture units then a 4890 = Win

Doesn't matter about the ROPS as they haven't been a performance bottleneck for some time on ATI cards now.

When overclocked it should be very fast and right behind the 5850, This could be the card to get this generation

 

Quote:

Originally Posted by antuk15 So it should be about 40-45% faster then a 4890

i'm thinking 20%

 

Cypress performance over rv790 is so much lower than you would expect by counting the Sp's. And the big problem is not bandwidth it seems.

 

cant wait to get my hands into some of these

 

HD5830 has higher TDP than its bigger brother HD5850 ! Also, in some aspects it is worse than HD5770 (like pixel fillrate for example)

 

Quote:

Originally Posted by antuk15 So it should be about 40-45% faster then a 4890

Even HD5850 is not 40-45% faster than HD4890 on average. Not even close to that
http://www.anandtech.com/video/showdoc.aspx?i=3650&p=5

Quote:

More texture units then a 4890 = Win

It has slightly more texture units, but also slightly lower clock speed

I fail to see how a video card with slightly higher texture fillrate would perform %40-%45.

Even %20 is an exaggeration

Quote:

Doesn't matter about the ROPS as they haven't been a performance bottleneck for some time on ATI cards now.

How do you know that? evidence ?