New 2012 Models Speculation Thread - PxxxEM Series (P151EM, P150EM, P170EM)

So basically the 675m should be the newer 580/485m, I hope it can match grx560ti stock clocks now; at 900mhz/2100 the desktop 560ti is more than enough for everything.
I'm curious about the 680m thought.
Anyway the waiting is killing me, we need some real proof and data on next gen nvidia gpus..Im really looking forward to step up from my 485m althought I'm extremely pleased of its performance..

 

No way in hell that the GTX 670M is going to be 100W TDP. I just can`t believe it. If anything, the heat output by these cards should go down compared to Fermi now that we got a huge die shrink. Plus the hotclocks are gone with Kepler, leaving the shaders operating at a lower frequency. People are whispering that the Kepler will be more effecient clock for clock too, Nvidia have shown in slides that the performance/watt have gone up

There has to be something wrong with the info from Monster. The 570M was around 80W TDP, with the above in mind, why should the 670M go up by 20W? And where does it leave 680M? 130W? And why are 670M and 675M both 100W?

 

I'll literally kill myself if you were right and the 680M actually might turn out to be 130W, imagine the freaking size on the charger, my god..

 

you mean figuratively kill yourself jaug1337

 

If it turns out to be 130w the brick will be big enough to kill him lol.

 

Do any of you think that the P150EM will ever be able to be fitted with a 120HZ/3D screen, this would be ideal for me! I am thinking about going with the P170HM/EM at Malibal just for this option.

 

I remember 3D screen requires different no. of pin for the connector

 

p150em


nvidia the way it's meant to be renamed

 

Hmm looks like the 570M, but less rops and lower memory clock.

The CPU in that is still sandy bridge?

 

As long as the 675M is roughly equal to a 560Ti I have a new portable desktop replacement, 675m+optimus = Power + Battery Life. Best of Both worlds.

 

The 580m right now is basically equal to a 560ti. It's got the same number of shaders/cores, just running at downclocked core and memory frequencies. Overclocking it can bring it in line with the 560ti in terms of performance.

 

nope.........according the tester, their gpu code:
GTX 670M=N13E-GS1-LP
GTX 675M=N13E-GS1

so i bet 675m will be renamed 580m
(bcz "lp" may be low profile, and 570m=trimmed down 580m)

 

This is bullcrap. If 670M is indeed a rebadge, then what is Kepler? 675M? 660M? Why a Fermi in the mix?

I somehow doubt the name. Well, I WANT to doubt it :/

Anyhow, here are the results from 3DMark with the "670M" if it is not a fake



Here is the source: Please post the source whenever you are posting these pics.
http://www.ocnb.com.cn/showtopic-11032.aspx

 

The specs may be similar but who knows what nVidia did with the Kepler architecture.

 

Look at the scores. 570M scores 17500ish on 3DMark06. 670M scores roughly 1000 more. And that is because of the slightly higher clocks

EDIT: No wait, 570M scores 18241 with 2760QM. 400 more with 670M....Its not even worthy to be a rebadge.

 

Kepler doesn't have hot clocks (shaders running at a higher frequency than the core), so I can say with 100% certainty that the GTX 670M is a Fermi chip.

 

wow. looks like im forking out over $600 for a Kepler then

 

If that really is the final version of the 670M, all that's worth waiting for are the 256-bit GTX 6XXM GPU's.

 

Yeah and pray that the 675M is Kepler and not the GTX 580M.

 

Let us also pray that the 680M is double the perf of 580M

 

Did anyone notice there's an extra |\ key next to the spacebar in that picture?

 

Why does that say the P150EM is 17.3"? I hate Chinese sources.

Edit: Just confirmed as fake, look at Device ID's on the 570m/670m, they're the exact same, 460m and 560m have different device ID's but they're the same GPU.

Or I could be having a moment, i swear that was a 1210 before. Someone needs to see if that's a photoshopped gpu-z to begin with.

 

Must be a typo. It is AUO17ED 15.3" screen.

570M:10DE-1210
580M:10DE-1211
670M:10DE-1213

 

Why is the 560m named higher than the 580m? 10DE-1251. And the 460m was 10DE-0DD1.

Seems very odd they would only increase it from 1210 to 1213. Something seems fishy about this whole 'leak'. The keyboard (that extra key) seems odd, the fact it's a Sandy Bridge machine...things don't add up, we have to wait a little over a week still for CeBIT though And I'm sleepy, but I swear they were both 1210 when i looked at the screenshots.

 

I doubt that picture. Look at the keyboard layout, since when they turn back non chiclet ?

 

Look at the youtube video of Clevo's upcoming model showcase, they are actually non-chiclet and look like what's in the picture.

That is an accurate photo.


Sent from my iPhone with Tapatalk

 

One other niggling detail, normal gpu-z shots show the 570m and CUDA/DirectCompute are checked, why are they not checked on that 570m/670m shot?

 

guys ive heard from a reliable source that optimus cant be implented with high end cards due to restrictions it brings up when implented!

dont get your hopes up but cards above the 650m wont have optimus!

i will be posting pictures that explain the way optimus works later and why it cant work

 

You do realize that the 580m for example, has Optimus enabled, just nobody implements it, because of the whole 1x link thing. Clevo has stated that the P150EM is Optimus enabled. And unless something drastically changes, the 650m won't even be offered as an option. Ivy Bridge brings more pcie lanes, and such, we'll see if there's not a new version of optimus for ivy+keplar. Current gen you lose performance yes, we'll see what Clevo has in mind for us with the P150EM.

 

i didnt say that the 650m will be implented in the 15 inch clevo ! maybe it will be implented in the 11 inch clevo or the lenovo y480

 

Does not Alienware M17x support Optimus with GTX 580M ? I think so....

 

you can check them yourself with a mouseclick

 

Nope, they were different IDs all the time. 1210 for 570M and 1213 for 670M. But I`m pretty shure that it is easy to spoof hardware ID of a GPU just as it is with a HDD or any other hardware.
I just don`t buy it. 670M a Fermi, while 660M and 675M a Kepler? And 460M and 560M were both Fermi, i see no point in continuing this slightly higher clocks for another digit up. As if the naming thingy Nvidia got going on isn`t confusing enough already

 

Well it has a Sandy Bridge processor, why wouldn't they spoof that too?

 

Guys the italian clevo builder Santech.it has the new clevos up for configure and ready to ship the 19th of March.
They are offering only one gpu thought: gtx675m 2gb ddr5.
So I further investigated on it and a reliable source told me that its a rebadged gtx580m running at 775mhz, but I wasn't able to tell me if it's Kepler yet.
The price should explain the rebadging since its priced 390,00€.

 

lol lol lol lol looooool

Seriously?! So the whole 600 series is rebadged 500 series? So the Ivy Bridge notebooks that have been revealed thus far, Alienware M18X R2/M17X R4, Sager laptops, MSI with their next laptops with 670M/675M, Asus G75 with 670M, will all be a pathetic excuse for an upgrade?
What about the early leaks from semiaccurate with good Vantage scores for the 600 series, all a bluff?


[/QUOTE]

?????

 

Maybe there is a 700 series lol. I think the entire 600 series is a rebadgings, but the entry Level 600 series are 40nm rebadgings, and upward are 28nm rebadgings, the 700 series is all Kepler 28nm. The reason behind this is to be competitive with the Radeon 7000 series naming 7xx?x. If I'm right I called it .

 

This reminds me of my Acer, it originally had Radeon HD mobility 5650, but later appeared with a rebadged version 6550m.

 

I just can't believe this is going to be true. It would mean that Nvidia has been misleading the media since CES, because the high-end GTX 600M have been called Kepler by numerous reputable tech sites, and Nvidia has never corrected any one of them.

Then there's that Nvidia has been pretending to have "samples", as if they were new prototypes.

It would be comical, and they're handing the mobile gaming market to AMD for 2012, at least until they get real Kepler mobiles released.

Just... no way.

 

But numerous sites have claimed that Nvidia was starting to ship Kepler mobile GPUs to notebook OEMs already in January for a April launch.

The 700M series have never been mentioned

But if the 660M, 670M and 680M are 28nm but Fermi, why on bloody earth don`t they shove in MOAR cores when they have much bigger thermal headroom? The 670M have the exact same number of cores as 570M. Have this happend earlier with GPUs? A die shrink, but same number of cores, almost the exact same frequency on cores and memory, same amount of memory

 

Ok heres some more informations based on santech website:


VGA:
SWITCHABLE GRAPHIC, NVIDIA OPTIMUS TECHNOLOGY:

NVIDIA GeForce GTX 670M 1,5GB GDDR5 dedicati
Memory Interface 192bit
Socket MXM 3.0 TypeB
NVIDIA PhysX, NVIDIA CUDA technologies
HDMI 1.4a - DirectX 11 compatibile

NVIDIA GeForce GTX 675M 2,0GB GDDR5 dedicati
Memory Interface 256bit
Socket MXM 3.0 TypeB
NVIDIA PhysX, NVIDIA CUDA technologies
HDMI 1.4a - DirectX 11 compatibile

Intel HD Graphics 4000
fino a 1.748Mb DDR3 condivisi
DirectX 11 compatibile, Shader 5.0


Tecnologia: Intel® CoreTM Mobile - Chief River (Sandy/Ivy Bridge 22nm), FSB 1.600MHz, 64bit
CPU:
Intel Core i7: i7-2760QM - 2,40 GHz - Quad Core - Cache 6 MB - TDP 45W - Turbo Mode up to 3,50GHz
Intel Core i7: i7-2860QM - 2,50 GHz - Quad Core - Cache 8 MB - TDP 45W - Turbo Mode up to 3,60GHz
Intel Core i7: i7-2960XM - 2,70 GHz - Quad Core - Cache 8 MB - TDP 55W - Turbo Mode up to 3,70GHz

Intel Core i7: i7-3720QM - 2,60 GHz - Quad Core - Cache 6 MB - TDP 45W - Turbo Mode up to 3,60GHz
Intel Core i7: i7-3820QM - 2,70 GHz - Quad Core - Cache 8 MB - TDP 45W - Turbo Mode up to 3,70GHz
Intel Core i7: i7-3920XM - 2,90 GHz - Quad Core - Cache 8 MB - TDP 55W - Turbo Mode up to 3,80GHz

So based on my understandings, Panther point chief river will be sandy and ivy bridge compatible. That means that the socket is the same if i understand correctly; intel could do have the same approach as desktop ivy bridges. Besides the dedicated chipset and motherboard, the current sandy motherboards could be compatible with newer cpus, or just more easily the newer panther point platform can be backward compatible.
What matters here is, if nothing changes, is that the socket is the same. If we can upgrade to a better cpu with out current machines it would be great

Also based on the source i was referring to on my previous posts, seems that those specs about the nvidia gpus are real and the gpu's are already in hand. Maybe the faboulus 680m will come sometimes later like july/august like it happened in the past, who knows. But so far, the 670m si infact a rebadged 570m and 675m is a rebadged 580m with higher clocks and optimus support.
Ive inquired them for some more infos, lets see what they say.

 

The 670M and 675m aren't even 28nm rebadges, still 40nm?

 

nah, they have to be 28 nm, otherwise the 675M wouldnt have a proposed TDP of 70-75W but a 100W one like the 580M.

cheers

 

If 675m is a 28m, process then I might just jump in on that since I don't think I can wait lol. I'll just upgrade later on.

 

I can't imagine they wasted time making 28nm cards which are identical to GTX 500M. It'd be insulting.

 

So the 670m screenshot of the GPU specs and 3d06 benchmark are just spoofs?

from another thread on NBR http://www.bestgaminglaptop.net/news/nvidia-28nm-mobile-gpu-plans-revealed/

 

What happen to the ivy quad core with tdp 35w? I swear I saw Intel mentioned it somewhere. With 35W on quad core, heat and power consumption are much more efficient.

 

OEM only, 3612QM.

 

And i doubt it'll be installed on Sager laptops, probably 45W 3610QM should be the lowest available quad core CPU!

I personally like the new look of a backlit keyboard, even if it's not chiclet! Can't beleive that 670M/675M are not Kepler GPU's, not gonna happen (or am i naive?)... probably just rumors, but I get more excited too bad we have no word about upcoming AMD HD 7970M just yet

 

NVIDIA Kepler Power Circuitry Revealed: 300W TDP? by VR-Zone.com

bad news?

My guess: 28nm rehash of the fermi chip, but inefficient

A fairly reliable source from semiaccurate:

SemiAccurate Forums - View Single Post - Any news on Kepler?