People seem to misunderstand what the gtx 485m is meant to be

I get comments on my YT videos for DX11 games asking why i'm running games like Crysis 2 (dx11), Metro 2033, ect. at 720p rather than 1080p.

People think it's capable of running all games at 1080p fluidly but that isn't always the case.

I consider a laptop with a single gtx 485m like a gaming console level of performance for DirectX 11 generation of games, like how xbox 360 runs most its games at a native res of 720p. Run DX11 games on this gpu at any res higher and it seems you only get sub 30 fps.

Now DX9 games or Xbox 360/PS3 multiplatform games, such as unreal engine 3 games (Batman AA, Bulletstorm) run great in 1080p as long as they don't use some high PhysX effects.

So my general rule when maxing out games on this gpu.

DX11 games = console gaming resolution (720p)
DX9 and Console ports = anywhere between 900p-1080p

 

I also hope with driver updates it will fix up how some DX11 games run sub 30 fps even at 720p. It's kinda annoying even when i max out Metro 2033 or Crysis 2 with DX11 and use 720p i get the occasional sub 30 fps drops. Though maybe that's just problem with the games themselves and not the gpu drivers.

 

The GTX485M is more or less performance equivalent of a desktop GTX 460. Likewise the GTX 460/560M is the same core architecture as a desktop GTS 450 and the 6970M is closer to a desktop HD 5770.

So you can understand easily why even the top end mobile chips are not going to be able to do everything at 1080p with max eyecandy. That said, I would rather reduce other settings than force myself to reduce native res. I find the blurry interpolation horrible.

 

I notice missing graphical effects more than a resolution drop.

 

It's literally a 6850.

 

is that really the case? i also prefer gaming at max resolution and ease back on the advanced settings. especially AA and AF are barely noticeable when comparing 2x with like 16x. ud actually have to use a magnifying glass to spot any differences but they eat up tons of computing power! the same goes for soft shadows, altho there the difference is a bit more noticeable. but compared to the "washed-out" look of interpolating a sub-native resolution its a totally good trade-off imho!

cheers

 

The 580M, according to Soviet Sunrise, kills the 485M outright. He has a card in his possession, and according to him, even with the same clocks, his games run better. He said he didn't understand it, we both guessed it was probably
1 - GF114 core > GF 104 core completely
2 - His 485M was from a bad batch
3 - Both of the above.

That's my two cents, take what you will from it. It's the only first-hand gamer experience I've seen.

 

I tried overclocking my 485m to the 580m speed and barely saw a difference in Crysis 2 dx11 at 1080p, maybe 2 fps difference. I think that guy is full of crap.

 

haha, nice one! rumor has it that the 5xxm series chips are optimized for tesselation and dx11 performance in general. altho i havent seen any hard proof for that yet to be honest. and looking at the ton of benchmark results already available on notebookcheck.com, ull just see a total of 8% (synthetic benchmarks) and 10.7% (game benches) improvement of the 580M over the 485M, respectively. those are pretty much exactly the OC values they implemented in core and shaders.... so yeah, lets wait and see what the 6990m brings to the table

cheers

 

You're a brave one to call Soviet Sunrise full of crap.

As for my post, let me fine-tune what I said. His 485M was overclocked. It was higher than the 580M's base clocks before the 580M came out. When he got the 580M, for testing purposes, he OC'd the 580M's clocks to match his old 485M's overclock. His games still performed better in general. So, as I ORIGINALLY POSTED, after telling him how the 5xxM cards were supposed to be more optimized for DX11 and tesselation, we came to the conclusion that:
1 - GF114 core > GF 104 core completely
2 - His 485M was from a bad batch
3 - Both of the above.

Also, I did not post this before, but when the clocks match, his 580M ran a good 5 degrees celcius cooler than his 485M did. This might be as a result of better thermal paste applying, but we can't say for sure. His 485M ran cool for him, so the extra temp drop was just icing on the cake.

 

ok, so....define "his games still performed better in general" in terms of fps, image quality, temps? this might all be true, but without specific numbers we cant really quantify this statement

cheers

 

Next time I see him online I'll ask for definite increases/gains, as well as temps. I don't see him *that* much on steam. He'll probably log in sometime this week and I'll let you all know the details. Steam doesn't have a chat log feature, sadly.

Though I can quote him: "I don't know if my 485M was from a bad batch or something, but man this card hauls so much *bleep*!"

EDIT: "bleep" being the prefix usually tacked onto "hole". I don't know if this forum lets me use that language.