Support NeoGAF

[marsomega]

http://www.anandtech.com/video/showdoc.aspx?i=2527

The 512-bit internal ring bus actually excites us the most; Cell's Elemental Interface Bus (EIB) uses a 16-byte wide quad ring bus, capable of transferring 16-bytes per cycle to the various SPEs. R520's ring bus would actually be four times wider than the bus found on Cell; though this implementation is in R520's memory controller rather than the PPE-to-SPE interface. With all the hype for R520 so far, everyone is banking on the memory controller to put X1800 ahead of G70.

http://www.hkepc.com/hwdb/r520firstlook-e.htm

As the card maker doesn't wish to reveal the R520 performance yet, but we can confirm that R520 at 24pp or 32pp will not be lower than that of the GeForce 7800GTX. Even though the 32pp and 24pp R520 mass production will face some difficulties, but the 16 pipelines R520 using the 90nm core will be clocked high to maintain a certain level of competitiveness. Apart from employing TSMC to make the 90nm R520 cores, UMC will be making them as well in the near future to make up the shortage problem and we can see more of the 32pp and 24pp R520 versions.

EDIT:......

http://www.anandtech.com/video/showdoc.aspx?i=2532

Pricing!

X1800 XT 16 600MHz 700MHz 512MB GDDR3 $599
X1800 XL 16 550MHz 625MHz 512MB GDDR3 $499
X1800 Pro 16 500MHz 500MHz 256MB GDDR3 $449
X1800 LE 12 450MHz 450MHZ 256MB GDDR3 $349

Common features to all R520 based boards include the new 90nm lead free manufacturing process, a Xilleon based TV encoder, SM3.0, H.264 decode acceleration and CrossFire support.

 

[Slurpy]

Hmm.

This may be my next card once in drops in price. Nice upgrade over my 9600XT.

 

[Pimpwerx]

Very nice card, but a few things.

EIB > this ring bus (at least @ 650MHz). EIB is 16B word, or 128bits. This bus is 512bits. That's a 4x increase in width, but Cell's EIB has a 5x advantage in clock speed (I assume it runs @ 3.2GHz, but please correct me if I'm wrong). So it's close, but not quite.

But this isn't a bad thing, since this a friggin GPU, and that sort of internal bandwidth is still huge:
64B * 650MHz = 41.6GB/s. Err...I think I got that right.

Another thing is that it's kinda disappointing to hear that the 16pp version won't necessarily beat the GTX. I mean, it's a 50% increase in clock versus a 50% increase in pipes for the GTX. It should be a draw, so maybe NVidia has more efficient pipes or something. Or maybe I won't say something so silly until I see a damn benchmark.

But I really hope the 24pp and 32pp versions come in clocked at that same speed or higher, and that they blow the doors off the GTX. Force NVidia to reveal the Ultra, at least, and drop the GTX and 16pp ATI card down in price to something more realistic. I also wonder how the memory bandwidth will factor into all of this. If shaders are long enough, it won't be a big deal, but it might not show the advantages it deserves running current games. :?

Anyway, the GPU wars are awesome. I repeat my mantra again, there should be no fanboys in this war. The only thing that matters is performance. ATI and NVidia are putting on a hell of a show the last few years. PEACE.

 

[antipode]

The 512-bit internal ring bus actually excites us the most; Cell's Elemental Interface Bus (EIB) uses a 16-byte wide quad ring bus, capable of transferring 16-bytes per cycle to the various SPEs. R520's ring bus would actually be four times wider than the bus found on Cell

This is wrong. The EIB is 4 16-byte wide rings (512-bit) and is capable of 96 bytes per cycle.

Edit - beaten.

 

[Pimpwerx]

This is wrong. The EIB is 4 16-byte wide rings (512-bit) and is capable of 96 bytes per cycle.

Edit - beaten.

No, you actually have it right. I forgot it was 96B and not just 16B. PEACE.

 

[isamu]

when will the first R520 enabled card be on shelves?

 

[xexex]

regardless of if the 512-bit ring bus is true or not, the R520 / X1800 GPU is still connected to memory via a 256-bit bus, like all other highend GPUs.

 

[Hooker]

^ FSAA

 

[hukasmokincaterpillar]

I havent been keeping up with the gpu wars since I got my GT, has VGA gone and died on me or something? There only be two DVI and one S-Video out on that there board.

 

[blackadde]

Will probably ship with a DVI->VGA adapter.

 

[XMonkey]

Heh, I see ATI's taken a cue from Matrox and arranged the RAM chips like a circle around the GPU (Matrox did this with the Parhelia IIRC).

 

[MrPing1000]

I havent been keeping up with the gpu wars since I got my GT, has VGA gone and died on me or something? There only be two DVI and one S-Video out on that there board.

they've been supplying cards that don't have it with a little adapter these days.

 

[Hooker]

http://www.vr-zone.com/?i=2666&s=1


The x1800 XT scores 8600 in 3D Mark 05... that's a LOT!

edit: Ah, they've taken it down already :/

 

[marsomega]

regardless of if the 512-bit ring bus is true or not, the R520 / X1800 GPU is still connected to memory via a 256-bit bus, like all other highend GPUs.

regardless of if the 512-bit ring bus is true or not, the R520 / X1800 GPU is still connected to memory via a 256-bit bus, like all other highend GPUs.

Exactly, also there is the A-AA and the AVIVO which the article couldn't go into because of NDA's. They both have extra A's so it has to be good... :lol

 

[marsomega]

Updated with more details (pricing.)

And more interesting details.


http://www.anandtech.com/video/showdoc.aspx?i=2532

At the top end is the Radeon X1800 XT; this 16-pipe R520 will feature a 600MHz core clock, with a 256-bit memory bus connected to 512MB of GDDR3 memory clocked at 700MHz. The 600MHz core clock will give it a lower fill rate than the GeForce 7800 GTX (24-pipes at 430MHz), while the 700MHz memory clock will give it more memory bandwidth than the stock GTX (600MHz). Much like the GTX, the X1800 XT will be priced at $599. The X1800 XT will feature two DVI outputs with HDCP support. The lower fillrate seems alarming at first, but consider several factors. First of all, ATI's traditional core design can do "more" per clock cycle (at least on the R420 design) than NVIDIA. Secondly, R520 has a lot of little tweaks including hardware asissted H.264 decoding. Just last week, we also received details about ATI's revamped memory controller which operates on an internal 512-bit ring bus. There is a lot to speculate about performance, but even with similar fill rates as NVIDIA, there is a strong possibility that other workings in R520 will differentiate the card on a real world performance level.

Next up is the Radeon X1800 XL, which is positioned between the GeForce 7800 GTX and the 7800 GT. The XL drops the core clock down to 550MHz, and the memory clock down to 625MHz. Other than the lower clock speeds, the XL is identical to the XT, meaning it still has 512MB of GDDR3 memory connected to a 256-bit memory bus. The X1800 XL will be priced at $499. Both the X1800 XT and X1800 XL appear to be dual-slot designs from previous roadmaps and existing box art. The roadmap also details that there will be HDCP support for the X1800 XL and X1800 XT via Texas Instrument's TFP513PAP DVI transmitter.

Priced at $449, we have the X1800 Pro, once more a 16-pipe R520 design but this time the core runs at 500MHz. The Radeon X1800 Pro only has 256MB of memory, also running at 500MHz, but still retains the same 256-bit memory bus. What is interesting about the Radeon X1800 Pro is that its fill rate and memory bandwidth appear to be identical to that of NVIDIA's GeForce 7800GT; coincidentally, so does its price. The reference design for the X1800 Pro features a single VGA and a single DVI connector, with no HDCP support.

The last member of the R520 family is the Radeon X1800 LE, which disables four of the pipelines of the R520 taking it down to a 12-pipe design. The LE runs at 450MHz with 256MB of 450MHz GDDR3 memory. Once again we're dealing with a 256-bit memory bus, and this time a $349 price tag. The outputs are identical to the X1800 Pro. Both the Pro and LE cards are single slot cooling design, thanks to their lower running clock speeds.

Hmmmmmmm......The yield problems were just as bad as they were true? Looks like ATI's magic new rev to improve yields didn't help as much as they claimed..