NV30 Architecture Info

[rdytorave]

http://www.extremetech.com/article2/0,3973,710350,00.asp

The biggest point in the article to me was that the NV30 won't be coming out in large numbers until February.

 

[Pinky]

The biggest point in the article to me was that the NV30 won't be coming out in large numbers until February.

Fine with me, I won't be upgrading for a long while anyways, and by then the prices will be pretty schweet .

 

[Rav]

I thought it was funny that the card only has 16GB/s of memory bandwidth. NVidia claims they have a lossless 4:1 compression technique, but I think thats being awfully optomisitc, and we'll be lucky to see 2:1 in actual situations. They should have used a 256bit bus and cheaper DDR. this 1Ghz DDR deal is just marketing. The cooler on that thing is huge, and I guarentee its very loud. Also, their Perot-esque charts show us big numbers that are overkill even a year from now. I rather have improved technology for whats out now than support for technology that won't be out until the card is obsolete.

-Rav

 

[quegyboe]

I find this all hard to believe. Nvidia is claiming to use DDR2 on the NV30, but they claim that DDR2 won't be widley available until Q3 2003, which means that this card won't truely be available until Q3 2003.
Also, alot of the stuff in that article seemed to be alot of tech talk, but I didn't see alot of mind blowing new tech. Sure they have the 32 128bit floating point processors, but they are still using a 128bit bus and things like that.
So far, from what I have just read, I'm not all that impressed. The NV30 still uses Z buffering, (no TBR), and still has a 128bit memory bus. Nothing amazing, I was more impressed with the Radeon 9700...

 

[OC Noob]

500+Mhz Core clock speed?


Now I think they are blowing smoke up our a#$es.


What a joke.


I wish they'd quit crying for attention and put the darn thing on store shelves.

 

[Top Hat Theater]

Or we could all be pleasantly surprised and they make good on all their promises. But then I'm just bein' a fool

~THT

 

[Rav]

I think we will get whats expected of a card that comes out 4-6 months after 9700 level technology. I'm looking forward to seeing ATI's responce too.

-Rav

 

[Wackygore]

I was hoping for some form of Tile-based rendering (TBR) to make up for the NV30s lack of raw bandwidth. With out it I'd have to put my money on ATI next year. I just can't see Nvidia's compression technique besting an R300 with DDR2 next year.

 

[speedy4500]

From the pics I've seen on the net of the GeForce FX cooler, it looks to be not very effective. (Scroll down on this website to see an example). There has got to be a better way to cool it. I think that the cooler is another suggestive marketing scheme...big cooler = big heat = big performance.....? Not always true. Look at the thing. The intake for the fan is directly next to the exhaust...it will be pulling in warm air from the exhaust to heat it up even more, and send it out to be taken in again. They'd be wiser just having the intake inside the case, pointing towards the front, don't you think? And I don't see how the heatpipe is more effective, it's just moving the heat to another place to be cooled by air. It's like making Jello and instead of putting in your refrigerator, taking it to your neighbors refrigerator to cool, you're not accomplishing anything spectacular at all.....we've read Joe's heatpipe reviews on the front page, and for the most part, they're not nearly as effective as a simpler well engineered air cooled heatsink. How can nVidia's chips produce so much heat and need so much power that they need the big heatsinks and extra power circuitry that makes the boards so big, when ATi only needs a shrimpy aluminum heatsink and fan, and a PCB the size of a GeForce2? I'm gonna have to say marketing BS on this one.