Recent content by Shadowmage

I agree with what you said, but I do believe that NVIDIA does quite a bit of custom logic (shaders) as well as a lot of manual place and route. Since they already have 40nm GT21x coming up in the pipeline, maybe they considered doing the straight shrink but determined that it would take too long...

I think that they were just lazy because of no competition :p

That's a very good point. Architecturally, Larrabee isn't very efficient due to the fact that it's very general purpose. However, I bet that Intel can meet or surpass ATI's RV770 transistor density easily.

Also, NVIDIA's GT200 has a very very low density (think non-fill-cell-area/total-area), so NV can do much better in terms of perf/mm^2, etc. I would think that something with a very optimal, regular layout can achieve much much more (like RV770, sorry for the low blow NV :P )

I'm echoing the sentiments of the other people posting in this thread: 4870 in the worst case, most pro-NV reviews, ties the GTX 260 overall, and in the best case, beats the GTX 260 by around 5%. Here's the Anandtech review, for example: http://www.anandtech.com/video/showdoc.aspx?i=3341&p=13...

The point of the extra TPC is to beat ATI's 4870, which NV's current GTX260 can not accomplish. Of course the performance difference is something to "shout at", or else NV wouldn't bother doing so: after all, disabling 2 TPCs will increase yields compared to only disabling one.

Is there a RV770 article in the works? This is definitely one of the more interesting architectures in recent history and definitely deserves an in-depth B3D article. Hopefully you guys have the time for all this.

In addition, GT200b gx2 boards will still be much much more expensive to manufacture than RV770x2 boards, due to die size, power delivery, and bus width. It's always possible to make a higher performing graphics card: the question is always, is it worth the additional cost.

What NVIDIA needs now is the new equivalent of G80->G92: a die shrink with half the bus width, GDDR5, 55nm, and much higher clock speeds. However I still think that RV770 would beat this at high AA. On a related note, do we know yet how RV770 performs so well at 4xAA and 8xAA? Do you think it...

Could you expand a bit on what you mean by "area optimization"? Was the optimization at an architectural level (as in changing the function and structure of units) or at a synthesis level (making the synthesizer choose smaller cells, improve area utilization and routing, using custom standard...

My guess is that they did the shaders with custom layout, or perhaps.... Fast14? :wink:

It's finally at Newegg at a reasonable price! $270 http://www.newegg.com/Product/ProductList.asp?DEPA=0&type=&Description=x1900gt&Submit=ENE&Ntk=all&N=0&minPrice=&maxPrice=&Go.x=0&Go.y=0

I don't know much about the technology behind it, but I thought the Golden Sun series looked awesome.

more advanced degree, any time.

Let's not forget that the R580 features larger Z-buffers as well, which is also in the transistor count. EDITED for typo