• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

WARNING: S|A - Nvidia's GTX480 case dissected

I'd want that kinda cooling too if I was running three cards in multi GPU that close together even if they were just 9800GTs, specially if its running 24x7 in promotional situations.
 
I dunno about a fail, its a bit of a weak effort IMO tho not half as bad as some people are making out. For me the launch will be something of a non-event as I don't really care - tho it would be amusing to see some of the kiddies 180 on allegiances if it did trounce the 5 series.
 
I really hope it does trounce the 5 series. I have never been able to get on with ATI cards. i alwasy seem to have some problem or another but i may be forced to go red this spring if Fermi holds no competition/price comparison
 
I shouldnt want the 480 to fail as a market dominated by one company is never a good thing yet for some reason part of me wants to see it flop. :confused:
 
Jono8 said:
I really hope it does trounce the 5 series. I have never been able to get on with ATI cards. i alwasy seem to have some problem or another but i may be forced to go red this spring if Fermi holds no competition/price comparison
Click to expand...

The bad thing is, the closer it gets on prices to the AMD cards, the worse it is for Nvidia. Honestly AMD can make a killing on a 5870 and a minor profit on a 5850, while I'm fairly sure a 470gtx at £300 would make a loss. Generally speaking AMD/Nvidia are wanting to make cores for $50-100 a pop on the high end cards and make a good 50-100% profit on the cores, thats profit ignoring R&D and other things like marketing. Realistically I wouldn't be surprised if they need at least a 10-20% profit to cover R&D and everything else.


People like Rroff keep forgetting, in reality the 280/285gtx were quite easily faster than the 4870, the 4870 got close often enough and offered more than enough peformance for most. Even the price/performance wasn't the key factor to it doing so well, it was the fact it was £200, like the 5850 which is also under ridiculous demand.

THe 480gtx could be 10 times as fast as a 5970, but if it cost £800, it would sell less than a couple thousand worldwide. Nvidia really can't afford to just take 10's millions in losses on each months supply of Fermi's, no one really can.
 
Happy said:
Well it still looks to be a success from a Cuda point of view :)
Click to expand...

Nope, Nvidia kills its DP speed on desktop cards to around 1/4 of its full speed, to persaude people to shell out for a professional version.

THey can't produce enough to supply the demand in the professional market, or the gaming/home user market. FireGL will gain major marketshare in the professional market purely by being available. Already at Cebit companies that have never touched FireGL cards before and never shown them off, have them in all manner of computers because they can't get their hands on anything Nvidia because there aren't any cards.

It doesn't look good when Supermicro who don't show off firegl cards have two comps with cardboard Fermi's in, no specs and no availability date, hinting to people they don't think they'll have more than a handful of them available before the end of the year, while a bunch of their computers have real working FireGL cards in running demo's.

Its a truly awful advert for Nvidia that huge server guys are moving towards whats available thats not them. Considering Fermi's obscene cost per core and the massive losses they will take on desktop part sales, their only hope for profit from Fermi is the professional sector. But in general those guys won't order 200 Fermi's, or 20 Fermi's, they want to make a constantly suppliable product so would want to order 50k, or 100k of them. Without much supply its going to be hard to even sell Fermi to those guys in such small numbers.

Likewise for professional workstations price isn't always an issue, availability certainly is, power usage and heat generated most certainly are a problem and again this is all fairly bad news for Nvidia.
 
Last edited:
I meant from my point of view either of these cards will show a sizeable improvement over my GTX 260s for F@H capabilities.
 
That is some serious ducting for sure, the only up side i can see here is that 3 gtx480's will apparently run on a 1000w psu :D
 
"On September 25, 2009, Vijay Pande revealed in his blog that a new third version of the GPU client was in development.[28] GPU3 will use OpenCL (preferred over DirectX 11's Compute Shaders) as the software interface, which may mean that the GPU core will be unified for both ATI and nVidia, and may also mean the addition of support for other platforms with OpenCL support."

http://en.wikipedia.org/wiki/Folding@home

5870 = 2.72 TFlops
Fermi = 1.05 TFlops

Does this mean 5870 will be better at F@H with it's higher SP shader performance?
 
You must log in or register to reply here.
Back
Top Bottom