HURRAY: Nvidia 600 series not just Fermi!! (Kepler)

[Meaker]

performance/price amd was mostly (maybe not always) better anyway, but still I disagree with Fermi was worse than 5000m, it was MUCH better in performance, much more costly, but think the following please

gtx 460m = ati 5870m, 460m is about the coldest running GPU on planet, while 5870m hits about 90C easily.. it has maybe %0-5 better performance than 460m, but 460m can overclock like hell.. AND I am sure nobody will disagree on the driver support, nvidia is like 10 times better than amd in that case (I had ati's prior to my 460m, and NONE of the drivers installed fluently...) AND THEIR PRICES ARE THE SAME! 485m is about 40% better than 460m, definitely it is going to be much more expensive...

The 5870M could overvolt and clock much higher than the 460M, a card that I might add released months after it.

Also I notice how you mention the 485M, but gloss over the 480M. The 485M launched a year later than it should have done.

[aduy]

the 485m should have been called the 580m, and then 580m>585m, theres barely any difference between the two other than a 7% overclock and lower core voltage.

[ryzeki]

Fermi was a disappointment??? dude 485m was like a million times better than 5870m, 400m family is like the best thing ever happened in terms of GPU performance (8000m family was also nice)...

Dude... 485m was not even competing against 5870m (competition is in the same price range... you could get 2 5870m for less than the price of 1 480m, and get much more performance from AMD). They released the 480m to win the performance crown in single gpu and it was a disappointing, overly expensive card that barely beat the HD5870m.

Fermi overall was pretty average. 460m was barely keeping on par with 5870m, 470m took forever to launch, 480m was a laughable attempt at high end. 485m is an amazing card tho.

But aside from 460m and 485m.... the whole 400m was basically a disaster. the 8000/9000 series was much better in terms of producs and performance they offered. G92 sure lasted a long while haha.

[Cloudfire]

Good that someone remembers the facts. I have forgotten about the specifics

[awakeN]

Any news about the 660M? That's the Kepler chip I've set my eyes on... any benchmarks or anything?

[plancy]

The way it's named, the 660m should perform under the 670m, which essentially is an OC'ed 570m.

[Torment78]

The 5870M could overvolt and clock much higher than the 460M, a card that I might add released months after it.

Also I notice how you mention the 485M, but gloss over the 480M. The 485M launched a year later than it should have done.

I agree 480m was the worst card ever i think i had two so i can speek
485m chould have been the 480m but they f up and used a GF100 that is Hot as hell and need power and shoulnot been put on a note book card for any reason

[awakeN]

The way it's named, the 660m should perform under the 670m, which essentially is an OC'ed 570m.

No, it's not. The 660m is Kepler, whereas 670M is Fermi. The name doesn't necessarily determine the performance.

[Kevin]

It took Nvidia three tries (480M -> 460M -> 470M), to create a decent 400M which could edge out the 5870M, and two of those were ridiculously expensive plus only in Clevos. That's why people were disappointed.

There was no card in the 400M series worth praising. And don't throw the 485M at me, because AMD simply yawned and equaled it with the 6970M, at like $350 less.

Now here we sit, today, and the 600M series is looking like a disaster. We already have our confusing rebrands, and our "brute force it" 480M equivalents, so there's only the GTX 680M left to keep the entire series from being a total farce.

Because this is all clearly just a stall tactic. Read it for what it is. Nvidia does not care about the 600M line, because just surviving until 28nm production is ready to pump out the refined 700Ms is their true goal.

[anycolour]

Any news about the 660M? That's the Kepler chip I've set my eyes on... any benchmarks or anything?

^This. Need benchmarks, all that good stuff. Thank you.