View Single Post
Old 03-14-10, 03:13 PM   #39
Registered User
Join Date: Jul 2003
Posts: 1,526
Default Re: Realistic time frame for Fermi (not happy with 5870)

Originally Posted by Iruwen View Post
Must of this simply isn't true, so I won't even bother to argue about it EDITED by MUYA Less of the calling out /goading please

/e: I'm sorry. So I'll argue instead Not about the real tesselation hardware thing though Obviously it isn't hot because the GTX 470 has a worst case TDP of 225W, also this could be seen (or heard and felt) at the Cebit. Just because the fan can draw 1,8A at max load, that doesn't mean it will. All GTX 400 at the Cebit have not been loud. The GTX 280 had a 0,48A fan - so it was way more silent that a 5870? Also, this doesn't have to be the reference design, I think like the final clocks it wasn't sure at this time. Since the real world performance isn't clear yet, nobody can say how much power they will consume in relation to their performance. If they use 10 or 20W more than an ATI card than that actually isn't much and one has to keep the cards' featureset in mind. The same is true for the power supplies (don't focus on power supply recommendations but rather on how many amps are required), a GTX 285 required 40A, the 5870 requires 40A, the GTX 480 seems to require 42A so the GTX 470 will be en par with the other cards - how good or bad this is again depends on real world performance. The final clock speeds aren't even known yet.

Though on a general principle,would they design the cooling and specifically the amps used by the fan motor,if there would never be a case where a fan that powerfull wouldn't be needed?....That's the main issue,as engineeirng teams are extremely competent,and don't add things just for the fun of it.

just my 0.02 cents.
shadow001 is offline   Reply With Quote