55nm GTX 260 uses MORE POWER & puts out MORE HEAT.

http://en.expreview.com/2008/12/27/the-first-review-of-55nm-geforce-gtx260.html

Please read the article entirely before you post.

I want to discuss this because if these results are correct then the GTX 260 55nm is a very pointless card. WTF???

Now I have read other places that the 55nm will use almost 75W less power, but those also were not quoting real world benchmarks.

Please discuss.

Thanks to DeadlyDevil666 for pointing this out.

 

Yes, i am very upset, i waited to upgrade till the 55nm came out as i was sure it would be better as i didn't wait for the 55nm 9800 GTX and it owned the 65nm one.. heh

It's all good i guess... I'm just going to get the XFX black edition

 

55nm vs 65nm simply by architecture can't use more power than it's 65nm brother, I don't think you should be basing everything off of that one site. Wait until we see some reviews from AnAndTech/XBitLabs/BitTech

Also the heat output is directly related to the power usage.

 

You didn't read the article.

They are assuming the heat output is because the heat sink on the 55nm is smaller then the 65nm.

I agree though that we need to wait for more tests, but it looks like they did a decent test on the thing.

 

According to that review it uses a smaller heatsink than the 65nm version, so it is possible but best thing to do would be to wait and see.

 

The 55nm 9800GTX+ only gave a benefit of out of the box stock overclocking. If you volt-mod or use after market cooling then you wouldn't see very much gain going to the 55nm G92 compared to the 65nm. I mean hell I have my 65nm G92 @ 870/2106/2000, now show me a 9800GTX+ running those speeds without the same volt-mod.

 

Hmmm the "benefit" of 55nm is starting to seem like a pipe dream.

 

Yeah, im strongly considering just going with that XFX card... lol :banana:

 

It depends on the architecture and the power usage. If you use a smaller manufacturing process and the same amount of power then you should be able to hit higher clock speeds.

Compare the 90nm G80 to the 65nm G92, the clock speeds on the G92 (65/55nm) are HUGE, so the decrease to 55nm produced a smaller jump from the already hugely popular overclocking 65nm G92. You can see how much a 25nm decrease helps in comparison to a 10nm decrease.

 

One Possibility is, EVGA released 3 cards very fast, a stock 576mhz, a 626mhz, and a 670mhz card, some of the overclocked cards use more gpu and ram voltage, it might be that they rushed the cards and used the 670mhz voltage (above stock) on all 3 cards.

 

My 2 month old 260_216 is sitting at 733/1500/2200. The temps never go higher than 62c. i.e. Fallout3> 84 to 112 fps. I have no artifacts in games. Stock fan@72%. Game at 1680/1050 4xAA all high.
I would like a good reason to step up....

I've temporarily removed the water block from the card to "step-up"

 

but this doesn't explain a "supposed" increase in power usage, power usage would increase heat, hence why that would make the article make sense, so even if it uses a smaller heatsink, who cares, it still produces less heat then the 65nm counterpart, as i do not and would not believe that the 55nm uses more power

 

no mate you have got that wrong.

 

have you even read the article??

I mean, yes in theory it is, and this article needs to be confirmed by the other hardware sites.

 

i read the article, maybe more heat since the smaller heatsink, but not more power.
then again, that heatsink is barely smaller at all.

i call bull****.

waiting for the guru3d review.

 

well the card was run 675 the card that they were comparing it to was run at 575,just the higher clocks alone would have the card running hotter.
the hs does not look too much smaller....
like said we need more info....

 

interesting! im looking forward to more reviews, and some "benchmarks".

 

Seems pointless to conduct a test with diffrent paramenters for both the cards, mind you the heat output from raising clocks is not linear, so as others have said I wouldn't put much faith into this test.

Now interesting would be a 65 nm 260 at 675 clocks vs the measured temps and wattage useage of the 55 nm one. Seems their whole test is fail.


/edit,

After reading the thread abit more carefully it seems both cards are tested at stock clocks ( same clocks ). Only advantage I can outread fromt he 55 nm process is higher overall maximum clocks, wich you pay for in more heat due to a in fact smaller heatsink with less flenses. ( Why didn't they just put the heatsinks on a scale to just prove it is less metal to conduct the heat instead of that SS :x )

I'm sure they will be able to sell this card to people sitting on older cards for sure, it is not a great upgrade, but then again neither was 9800 GTX+ over 9800. A good / luck example of a 65 nm 260 can still be clocked over 700 where a bad 55 nm might stay at 700 so not much of an upgrade, no

 

lol

it does run hotter expreview or something took it apart and compared the heatsinks with the 216 core and the 216 core had a more effcient heatsink.

 

I don't know why there isn't one yet.