Pic of GTX 480M + 3DM06 Test

lol
never owed one since it was a rebrand of the 9800m

far cry 2 when all the all the explosions near the end of the ranch long run
far cry 2 about 1 minute in to the ranch long bench when your curving down the road out in to the open country side.
devil may cry 4 3rd test when you fight the dragon
metro2033 when you start fighting the long armed dwelling monsters

come on now, how are you the only person in the world with no micro stuttering??
rotflmao

bet you had stuttering with that 260m you use to have, tell the truth.

lmao!

yep, pretty pricey as usual...



side note:
the 480 will still win in the end.
no need to argue, when they start hitting the market and after like 1 or 2 months, they will do just as the 5870's did to the 4870's..

but of course...im not card specific. so we shall see...but that's my prospective....

 

This almost makes the new gargantuan chassis pointless, until version 2.0 at least.

 

Were not saying its not better, but the huge price just cant be justified by the little performance gain, plus theres the higher heat and power consumption.
I am certain there will be people who will still buy it - nvidia fans and those who just want the bragging rights, but for those who just want a top performing gpu with nearly identical performance for a much lower price i think we have prooven which is the better pick.

 

How is being the same speed, running hotter at 3 times the price and needing a bigger PS wining in the end? Sounds more to me like losing at the beginning.

 

do you have a 480 in your hands?

not talking about the price performance stuff..i leave that up to you guys.
im not concerned about the heat it put outs. i have my own cooling mods.
i understand where you guys are coming from..nothing wrong with your points in the least bit. but if i benched 5870s already, then why would i stay with them? i got what i wanted now time to move on.(from a benching point of view)

again, refer to upper post.
all that matters to you who care about that sort of thing..but people who mod and do their own thing..find work around's.

ill let you guys talk about heat/price and performance is yet to be seen when over clocked. since that's what every one really wants to know..

guys, this is just an opinion of mine. not me stating any facts. if i were stating facts..i would be benching a 480m right now..
or referring to mandrake/scook9/joker5150/quadzilla/dondadah and the many other pioneers of benching.

 

I don't know about anyone else, but I am not convinced yet it will be only very minimal performance gain if any. I say wait a couple more weeks till ones have this card in their systems, like the W880CU and such. I want to see actual skilled users on this forum check the 480M out and report what they find. Also, when SLI is released next month for it, that will be very interesting...

 

that's all im saying right there!
skilled users!

 

Guess then we need a raise of hands for all the "skilled" users who are actually buying these machines ...


I believe there is a W880CU Pre Order thread but im not sure i saw anyone actually say they had it preordered...


Hemi hopefully you can get a loaner for a week or 3 to beat up on ....


Someone needs to ship a system to Johnksss then we will find out what it can really do ... He has the patience of a saint for benchmarking .. Something im sorely lacking in these days...

 

rotflmao!

all that bashing going on...i didn't even want to try and ask....lol

 

Notebookreview is already in possession of and doing their review of the Sager NP9280 with the nVIDIA GeForce GTX 480M.

So, there is one "Skilled user" anyway

 

Good to hear.. i really do wonder how far you can push it in the D900F... If it at least overclocks well that could help it justify that price a bit... We were just commenting with what we have - stock gtx 480m vs stock mobility 5870... Well have to wait to see if its capable of more..

 

+1!
the first one out of the gate with someone WE know. nice.

 

Send me a test system. I'll happily compare it against my M17x. I'd be completely impartial as well. I think I qualify as skilled enough.

.

I'll check metro and far cry when I get back home on Friday. I nearly finished metro with my old 4870 xfire setup and never saw microstuttering. Of course with the already abysmal performance it's possible I could have missed it with that game. Why would a single 260m have microstuttering? I never had the sli.

The 480m probably come out a bit faster overall like we've seen but it will be a far cry from "smoking" the 5870.

 

you do qualify sir!

"smoking" the 5870. <-- this comes from a better min fps.

if we are all running around at 120 fps and get into a battle and your fps drop to 20 and mine stays at 40 for the next 15 seconds....who do you think is going to be in the better position? regardless if you get 240 fps staring at the stars.

it's nothing personal, but gaming..your more concerned about you min. frame rate..not your max. the min is going to kill you every time..not the max or average

yeah, you got me beat on games...i only have like 50 or so games and demos.

the 260 was to give and idea of how it looks...i guess im just not explaining right or something, but every system i have had showed signs of micro stuttering in places. and you have to serious be looking to see it. along with screen tearing. (this is real easy to see - move the mouse real fast in all directions to see it. hard core gamers use vsync no matter what) but with the fps so low. most pass it off as something else.

ill try to make a video of it...
on the right side of the screen is a long list of stuttering game plays
http://www.youtube.com/watch?v=rKa2NZVVxVk

 

Please do us a favor, by testing the 480M's overclockability.

 

You are asking the wrong person. It is NOTEBOOKREVIEW that is doing the review.

 

Who is it that will be doing the review ? Gophn(not sure he is even around much anymore ?) or Chaz ?..

The other thing we need to be realistic about here is not everyone is going to overclock these cards so what truly matters at the end of the day is stock performance...

So even if these cards end up opening a hole in the space time continuum when OCed that wont reflect everyday performance...

24/7 hardcore overclocking on laptop video cards is just not realistic... Sure people do it and may never have problems but even the big benchers will tell you after they are done doing the crazy stuff it gets turned back to stock ...

Either way i still wait anxiously to see some real results from people we know ..

 

I agree with quad. Oc is important but not everyone does it so stock matters for most consumers. However for us as a select group, oc means a lot as well.

Ps typing replies with an iPhone is tedious.

Have we seen min fps differences yet? I agree if one drops to 20 and the other 40 that's substantial but I don't see that happening in many games outside a select few if at all. I'll have to watch the vid later since my iPhone connection is iffy in georgia.

 

If it can be safely oc'd then I'd like to see those numbers. People that are buying these machines are the kind of people that do overclock I would think.

 

id agree its saving grace may be a solid, very high OC 30+ %

 

You would be surprised how many people buy high end machines and will not touch an OC with a ten foot pole...

Overclocking does matter alot to me to but generally out of all the laptops i have owned(to many to count) i would rarely if ever overclock during long gaming sessions etc...

If you have to strap an air conditioner to your laptop so you can safely game and OC at the same time then umm im not sure what to say on that one lol ..

 

joker:
no one would know about the min fps till the new card came out. and looking at the test results..the 480 has a higher min in just about every test ran.

sort of like how ati didn't think they had a problem with AA/AF till the 480 came out 6 months later...now they have a AA/AF problem. i was thinking i was king of the world with my dual gpu card...till my single gpu card came by and past it like it was sitting still...at quite a few test. *jaw hitting floor* was my reaction...lol
ditch that iphone and get an evo..
also, use a wifi connection or do like we do for messages... speech to text

 

I completely agree with you Quad.

You can do a permanent OC like I did to my card, but it takes a lot of testing, stability and temperature monitoring. If you want to keep your laptop going strong for more than 2 years with the same OC, this will have its risks unless you can keep the temperatures down.

And here is another review of the 480M:
NVIDIA GeForce GTX 480M, Fastest Notebook GPU Yet - HotHardware (Via laptopvideo2go)

 

wow, well that was a glowing review.....blew away the 5870 there, and by alot.

 

They are also using a desktop CPU vs. a mobile one, which is not really fair. So I would say the difference is more around 20% on average, just as I said so many times before. Still, it's enough to make me get the 480M over the 5870MR.

 

its by far the best review yet.

Also, keep in mind the 5870 OC's well, compared to the above benchmarks my machine would be tied/ahead....I'm also assuming the 480 will OC a bit and take back the lead. Just my 2 cents.

 

Hey he can also run the 2 5870s @ 930Mhz core with no external cooling....he's got a special system.

 

yeah he does!
jaw was hitting the floor on that one as well.

edit:
and looking at the next set of reviews...
the AA doesn't even move. it would do that with me as well, but i found that i had to force AA through ccc or tray tools for it to take affect. forcing it through the game didn't do anything.

 

Moar rumors out of Fuad today again claim that the GF104 will be better at overclocking than GF100 because it lacks the L2 cache.

So I'll say it again if they had a 336 Core/192-bit GF104 GPU in the pipes it would have likely made for a much better GTX 480M.

 

I fully agree. They also said in a previous post that the GF104 actually has "384 SPs organized as 16 clusters of 24 SPs" and "has a 256-bit memory interface, but with one MC block disabled you get the 192-bit memory interface".

I was actually pretty sure that they were going to use a GF104 core as it made much more sense. But I was wrong...

 

I've heard so many core counts for GF104 that I don't know what to think it'll end up having, but the only real mobile GPU we've seen so far (that's likely GF104 based) is the pic posted by Kevin Jack of the N11E-GS from the ASUS VX7 with it's 192 cores/192-bit. That should mean that somewhere out there on the internet Nvidia's got an N11E-G T packing more than 192 cores but less than 352.


One thing that most of those rumors have in common is that the core counts do seem to be made up of clusters of 24 (or 16) instead of the clusters of 32 that GF100 uses. Smaller clusters would probably make the chip smaller, cooler(?), and easier to manufacture...which should also mean cheaper. On the other hand smaller clusters and no L2 cache sounds a bit like G200 architecture. :wink:

 

Actually I don't use external cooling so yeah I guess I did have a special system. Not everyone goes out and spends cash on an external AC just to cool a laptop down..that's actually pretty ridiculous and ineffecient. If I was giong to go hardcore with it, I'd have a custom heatsink machined for the M17x instead and add in higher cfm fans than to stick my laptop over an AC--anyone can do that.

I was looking at benthedogtrainers dual AC setup and couldn't help but laugh. Now take what that dude that built OctaTron or whatever, that's original or what Soviet does with his heasink mods. Both of those require skill, knowledge and work. I got the score I did in Vantage because I figured out a workaround for the M17x/5870 xfire combo that nobody else knows right now. If I had simply pointed it at the AC I'd still be stuck at 16k. Thus I don't need any external AC cooling.

Oh and I agree with Phinagle, GF104 with those specs would've made much more sense for the 480M. We need direct apples to apples comparisons for the 5870 vs 480 instead of using two different spec'd machines. HotHardware failed in that regard.

 

out of curiosity, what is your highest single gpu score now, with this new found..errr, ummm workaround.
and does it work with the 920xm?

 

Yep you definitely have a special system/secret...I've tested 4 different types of ATI 5870s in 3 different machines. Couldn't do past 900 without special cooling on any of them. Plus there is no heatsink mod/fan speed that can take place of refrigerated air with high OC'ing although it definitely helps in some situations.

 

You all seem to forget that they need proper drivers before these things will start showing their true power. I'm sure both the desktop and mobile fermi cards have a ton of untapped potential. nVidia's beta 257.15 (maybe .16) drivers showed 10-25% increases over a span of games with their GTX 480. When they get themselves in gear properly, I'm sure the mobile card will follow suit; because on paper a single 480m should skewer SLI 285m. Not to mention ATI uses GDDR5 as well, but with only a 128-bit mem bus where the 480m uses a 256-bit one... 2 and 3 FPS increases in some games over a mobile 5870 just scream "you ain't seen nothin yet..."

 

they just have to wait and see...

 

Why do people keep saying the 256-bit bus is going to make such a big difference when they have the memory at such a low speed? The bandwidth numbers are not that different Nvidia is running at 2.4Ghz for 76.8GB/s, and ATI is running at 4.0Ghz for 64GB/s. To put that in perspective the 285 is 65.28GB/s. They have lost most of the advantage of a 256-bit bus by their need to run it at a low frequency for power draw. Also due to the low core speeds more memory bandwidth would just be wasted anyway.

 

Yeah, I agree with johnkss. We just have to wait and see.

 

btw ......3dmark 06 dosent count when you are testing a desktop vs a laptop system does it ?

 

Can't tell if those tests were done with a mobile proc or not....I'm guessing not.

 

Quick question, back to the benchmarks.....for a game like bad company 2, how the heck are they benchmarking a game that doesn't have a synthetic benchmark present? Are they just running fraps and playing through single player and noting the avg frame rate?

 

Another article on Fud today about the 384SP/256bit GF104 (someone else can link it, I'm on my phone) and how we'll only see cut down versions of it at launch.

All really confusing, but I'm thinking Nvidia went with the GF100 because it was a product they could already get to market instead of a product still another couple of months away. Fuad also makes mention that in the future we might see a full fledged GF104 show up as a GTX 475 in desktops...and I say something similar for the mobile market would be a good idea for Nvidia. A GTX 485m.

A GF100b probably wouldn't be a bad idea either.

 

Phinagle for president of nvidia? ^^,

Anyways here is the link I think:
Fudzilla - Nvidia's GF104 is crippled from the start

 

Eurocom launches monster laptop with 5870 CrossFire - www.nordichardware.com


Hmmmm, Eurocom updating X8100 with 5870MR CF, well, it's Eurocom, but at least it means that it will happen. If it's going to work on X8100, quite sure X7200 will have the same option as well to cut down on the cost.

 

Fudzilla - Mainstream Fermi is cancelled

Fudzilla - New Fermi with 512 shaders possible

To me, if all that is true, then Nvidia has realized that GF100 Fermi is bad architecture but they're going to sell it to consumers anyway just to make something out of all the delays and cost from development.

 

While it is sad for us as consumers, it does make sense for Nvidia... wouldn't you try to do the same?

Anyway, this is good news and it means that the next generation of Nvidia mobile cards will be much better.

 

It does seem inappropriate IMO. But of course if it's more powerful, then there's probably someone who would want it. Then it's not really that they want to sell it anyway anymore, it's really something that some might want.

 

my card can do over 900 on air with only the system fan ... about 920 but still I got a nice card so it is possible

 

i'm capped out at 880 =-/

 

Same here, get driver reset above 880.
hey moo, what is your vBios on your card ?