Did everyone bite too soon on the GTX 480? (GTX 485)

[Frag Maniac]

I think the real comparisons will be the revised 28nm Fermis vs ATI's next offerings. Neither camp has had DX11 hardware out long enough to perfect it yet, so a lot of this fanboyism talk back and forth is neither here nor there really.

[BeepBeep2]

I think the real comparisons will be the revised 28nm Fermis vs ATI's next offerings. Neither camp has had DX11 hardware out long enough to perfect it yet, so a lot of this fanboyism talk back and forth is neither here nor there really.

ATI will probably be on the 7000-series by the time Nvidia gets a 28nm part out.

[Systemlord]

I think the real comparisons will be the revised 28nm Fermis vs ATI's next offerings. Neither camp has had DX11 hardware out long enough to perfect it yet, so a lot of this fanboyism talk back and forth is neither here nor there really.

A die shrink to 28nm might lower the power requirement, temperatures and greatly improve clock rates, but a 512SP Fermi isn't that far from the current Fermi 480SP out now.

[Rollo]

Technically you could say "Did everyone bite too soon on (any card ever produced)?" because there are always refreshes and new models in the wings.

I don't have any faith in "Kit Guru" from what I've seen so far, agree with the assessment they post fake news to steal web traffic. Probably some knucklehead's scam to earn money for weed and pizzas.

Thread seems like flamebait.

[Rollo]

I don't think so. Heat issue with Fermi is exaggerated. I never saw mine go over 90 while gaming/folding with fan speed controled by AB profile.
BTW I'm used to heat, can't have top performace products without heat. Now, on water cooling (non-extreme, res - pump - EK - EK-vga - rad - back to res) I never saw it go over 50.
I would try 512 cores BIOS if I could find.

Side fans make a BIG difference with the 480s. I had mine in a Cosmos 1010 with a single 120mm side fan, and 2 X 120mm top and front intake, ran Metro 2033 at 92-93C . (480 SLi)

Now I'm using a HAF932 with 230mm on the side and it runs Metro at 84-86C (with some spikes to 87C) at higher res. Needless to say, the $139 shipped for the HAF I consider money well spent.

[Systemlord]

Technically you could say "Did everyone bite too soon on (any card ever produced)?" because there are always refreshes and new models in the wings.

I don't have any faith in "Kit Guru" from what I've seen so far, agree with the assessment they post fake news to steal web traffic. Probably some knucklehead's scam to earn money for weed and pizzas.

Thread seems like flamebait.

If anything I bit too late coming from an 8800GTX, I skipped the G80 to G92 refresh 8800GT, 9800GTX, 9800GX2, GTX 280, GTX 285 and GTX 295. If you play "The Waiting Game" there will always be something better coming out at least a year after your purchase! Guess what, Fermi will be replaced by Fermi 2 and so on and so on and this cycle will NEVER end! At some point you have to jump in and enjoy what you have and until you feel like your hardware is no longer adequate for it's intended purpose it's time to upgrade.

I've never heard of "Kit Guru" until the interview was given, his forum did come across as fake and attention seaking, his name should be "Kid Guru".

- Systemlord

[LedHed]

http://www.xbitlabs.com/news/video/d...This_Year.html

[bro20000]

all well am happy with my 480Gtx till next year! Nice find mate!

[Chrono Detector]

I don't think NVIDIA would be able to present a new high end card soon too as well, look at Fermi, 6 months after ATI released their 58xx series and that is quite late.

[Frag Maniac]

I don't think NVIDIA would be able to present a new high end card soon too as well, look at Fermi, 6 months after ATI released their 58xx series and that is quite late.

Why is it when everyone says that they totally disregard the fact that it was a bigger leap in GPU architecture than anyone has ever undertaken?

You're also disregarding the fact that a new high end card could very well use the existing architecture but with a die shrink.

Where the problem entered in for the most part as I understand it, regarding the 6 mo delay, was in trying to get all 512 cores working on the existing die size.

Nether the GF104 or a GF100 at 28nm would present such problems.