AMD: R8xx Speculation

[CJ]

Maybe due to the density of RV670?
They certainly increased density quite a bit afterwards with RV770.


I should hope it was not R800, that demo wasn't exactly mind blowing and it was only getting what some would call "mediocre" performance of about 20-30FPS.

I've been told the FPS counter on that DX11 demo was falsified... they set it to a certain amount so no one could guesstimate the final performance... just like they turned the wafer so no one could say anything useful about the architecture.

So do not be misled by that 20-30 FPS thingy.

 

[BRiT]

I've been told the FPS counter on that DX11 demo was falsified...
So do not be misled by that 20-30 FPS thingy.

Those where 20-30 Uber FPS!

 

[keritto]

Umm, no? Lecithin brotha', do you take it?

wtf? Lecithin ... another AlexV's trash talk? Is that some new ATi GPU chip

Because I didn't feel like paying an extra $50 for a <1% increase in performance 99% of the time.

What bandwidth limitation?

Please don't use yellow journalism quoting.

I just stated thta it wouldn't be a bandwith starved just like iRV770 with obsolete GDDR5 3.6Gbps chips have too much BW nowadays with 5Gbps 192b we would get 4% more BW on new architechture with possibly improved memory caching like RV770 had over R600 and more threaded code that heavily use registers instead huge latency cache. Or you just playing an all seeing big bro' that just can't be wrong?

 

[ninelven]

I just stated thta it wouldn't be a bandwith starved just like iRV770 with obsolete GDDR5 3.6Gbps chips have too much BW nowadays with 5Gbps 192b we would get 4% more BW on new architechture with possibly improved memory caching like RV770 had over R600 and more threaded code that heavily use registers instead huge latency cache. Or you just playing an all seeing big bro' that just can't be wrong?

I'm guessing English isn't you first language. Sorry, but I can't really understand what you are trying to say. I'll guess you think being bandwidth limited is a terrible thing. I see being bandwidth limited as good engineering and maximizing performance per dollar. You say Rv770 is bandwidth starved, and I ask what offers better bang/buck?

 

[Arty]

I've been told the FPS counter on that DX11 demo was falsified... they set it to a certain amount so no one could guesstimate the final performance... just like they turned the wafer so no one could say anything useful about the architecture.

So do not be misled by that 20-30 FPS thingy.

Falsified as in deliberately under-rate the actual performance?

And if it really is 20-30 fps, it should be slightly visible in the video. In such case they may have been better off borking the actual performance instead of fudging the fps counter.

 

[LordEC911]

I've been told the FPS counter on that DX11 demo was falsified... they set it to a certain amount so no one could guesstimate the final performance... just like they turned the wafer so no one could say anything useful about the architecture.

So do not be misled by that 20-30 FPS thingy.

Interesting...
Nothing like keeping us completely in the dark and screwing with our minds.

 

[keritto]

You say Rv770 is bandwidth starved, and I ask what offers better bang/buck?

Sorry but do you have dysleksia or what?

... just like RV770 with obsolete GDDR5 3.6Gbps chips have too much BW ...

 

[ninelven]

Sorry but do you have dysleksia or what?

Sorry, your inability to form a coherent sentence with correct punctuation makes things difficult. I am afraid I can no longer entertain your inquiries. Done.

 

[AlexV]

Hah, too much BW. Interesting perspective thar.

 

[Scrat]

I modified an old image found somewhere on the web:



About the same perimeter as RV670.

 

[pjbliverpool]

Thats an interesting image. I had no idea G92b was a little smaller than RV770 on the same process.

I know RV770 is faster but it does go to show that NV still had a pretty damn effiicient architecture with G80/G92.

 

[no-X]

I wouldn't be that optimistic. G92b/GTS250 seems to be the part, which helps nVidia in these bad days more than GT200b, but compared to RV770/HD4850, it has some drawbacks. It needs to run at 100-120MHz higher core clock (ROPs domain) and about 100MHz faster memory clock to stay competitive performance-wise. And still lacks DX10.1/GDDR5/DP support. Despite it I agree, that performance isn't bad for a 20 months old GPU.

 

[Love_In_Rio]

That's an article from semi-accurate:

http://www.semiaccurate.com/2009/06/09/ati-evergreen-code-names-explained/

Also, the original video I linked, as posted by amdunprocessed, showed "EG BROADWAY":

http://forum.beyond3d.com/showpost.php?p=1298835&postcount=723
http://forum.beyond3d.com/showpost.php?p=1298862&postcount=727

So Broadway is a codename that needs explaining too.

Jawed

Broadway was Wii´s cpu codename ( Hollywood the gpu´s ). Could it be "Wii 2" graphics chip ?. It would suit quite well with its tiny meassure. There are rumours of a new Wii next year:

http://www.gamesindustry.biz/articles/project-natal-to-launch-as-wii-hd-spoiler-analyst

 

[pjbliverpool]

Broadway was Wii´s cpu codename ( Hollywood the gpu´s ). Could it be "Wii 2" graphics chip ?. It would suit quite well with its tiny meassure. There are rumours of a new Wii next year:

http://www.gamesindustry.biz/articles/project-natal-to-launch-as-wii-hd-spoiler-analyst

Lol, that would be quite a step up for Nintendo! I doubt it somehow though, I expect the Wii has many more years left in it yet.

 

[rjc]

Thats an interesting image. I had no idea G92b was a little smaller than RV770 on the same process.

Not sure that the picture is correct, think the author has used rough measurements from an early PcPer article. There is a thread about it here. That said due to volume nvidia has likely been paying less per wafer over the lifetime of the G92/G92b giving it an advantage.

 

[itaru]

Due lack of high res evergreen die pictures, here's a lowres one, next to high res RV770

Even though lowres, it's IMO clear that the evergreen has gone through major changes over RV770-style, there's 4 clear "partitions" of the chip, while in RV7xx there's one big pile in the center

eDRAM or big cache ?

 

[Silent_Buddha]

Or as CJ stated, that's the back of the chip.

Regards,
SB

 

[Kef]

www.nordichardware.com has some more info:

http://www.nordichardware.com/news,9486.html

"AMD showed live DirectX 11 hardware at Computex, although with rigged FPS gage to hide the actual performance. The hardware was performance level, not high-end,"

"AMD is very confident about its coming DirectX 11 lineup and even though it used the same performance level hardware, and not high-end, it claimed it was the fastest single-GPU graphics card on the market. They said that the upcoming launch will be even more of a surprise than when they shocked the world with the 800sp RV770 GPU, still vague but very intriguing. "

/Kef

 

[silent_guy]

Or as CJ stated, that's the back of the chip.

Unless AMD suddenly invented a revolutionary new way to process silicion, that's complete nonsense. The back side of a silicon wafer is completely uniform and featureless. (FWIW, I have couple of scrap wafers at home.)

 

[itaru]