What's new
  • Please do not post any links until you have 3 posts as they will automatically be rejected to prevent SPAM. Many words are also blocked due to being used in SPAM Messages. Thanks!

NVIDIA's GeForce 400M Mobile GPUs: 7 New Fermis Introduced

Status
Not open for further replies.

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
After being thoroughly beaten in the race towards DX11 GPU sales, NVIDIA is working feverishly to make up for lost time. Nonetheless, it has been five months since the GTX 480 and GTX 470 launched amid much fanfare and only in the last six weeks have we seen derivatives take some shape. The GF104 took the guise of the highly popular and perfectly priced GTX 460 1GB and 768MB cards but even as it launched many were wondering: where in the world are the mobile chips? Well, today we have the answer for you.

Even though the GTX 480M was launched some time ago, the mobile computing market still had to endure the endlessly-renamed 300M / 200M GPUs when it came to anything under ultra high-end graphics needs. Granted, these are still highly competitive products in their own right but with AMD’s top to bottom HD 5000-series breathing down their necks, even the 300M chips started looking a bit long in the tooth. A true refresh for NVIDIA’s entire stable of mobile graphics solutions was necessary and this time, there are no baby steps; their whole lineup is literally being replaced in one fell swoop. Not one but SEVEN mobile GPUs are being launched today and should be featured immediately in laptops from everyone from Dell to ASUS to Lenovo.

The low end cards like the 310M and 305M will stay in place but the GT and GTX 400 series will be taking over the lion’s share of the workload from now on. If the numerous discounts on NVIDIA-totting notebooks in the last month have raised questions, you now have your answer as to why prices were slashed.

This short article will be focused upon one central goal: to introduce you to the new NVIDIA graphics cards in preparation for a full comparison against the competition. So without any further preamble, let’s take a look at NVIDIA’s new introductions and what these new mobile GPUs bring to the table.

GF100-41.jpg
 
Last edited by a moderator:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
A New Lineup, Top to Bottom

A New Lineup, Top to Bottom


The GPU industry is nothing if not extremely cautious when it comes to releasing new products so it is quite rare to see an entire lineup of graphics cards being refreshed at the same time. NVIDIA has bucked this trend by shedding much of their 300-series and 200-series in one fell swoop. One of the more interesting aspects of this is the fact that some of these GPU cores have yet to make their way into the desktop space. We have seen GF100, GF104 and the soon-to-launch GF106 architectures but could we be seeing the first glimpse of the rumored GF108 as well?


The Enthusiast Lineup

400M-7.jpg

In the high end we have some interesting things going on regarding the overall cost and performance comparisons against ATI’s HD 5000-series. According to our contacts at several system integrators, NVIDIA is targeting the performance and the price of the GTX 460 towards ATI’s current single mobile GPU champ: the HD 5870M. Meanwhile, the GTX 480M competes price-wise against ATI’s high end Crossfire solutions and the GTX 470M offers performance values that should eclipse the GTX 460M by about 10%. The NVIDIA GPUs have this ability to destroy ATI’s high end mobile offerings from a performance standpoint simply because many of the HD 5000M-series’ cores were seriously cut down when compared to their desktop brethren. For example, the HD 5870M sports only one half the number of Stream Processors that a standard HD 5870 does.

When compared against parts from the previous generation, there is no denying the fact that NVIDIA’s high end mobile GPUs were due for a change. The GTX 285M and GTX 260M went the way of the Dodo long ago which left the enthusiast end of the spectrum high and dry while the GTS 360 never really saw widespread adoption.


Mainstream Parts Aplenty

400M-8.jpg

Launching a whole five cards into the mid range and mainstream brackets is no easy feat but once again NVIDIA is looking to make up for lost time in a big way. Starting at the “top” we have the GT 445M which is supposed to be the price / performance leader among these new mobile parts. With 144 cores and the possibility of higher-bandwidth GDDR5, it should be able to compete with the HD 5850M and HD 5770M depending on the configuration chosen.

The GT 435M, GT 425M and GT 420M are all basically the same card but with varying clock speeds and memory allotments which in effect allows for a wider range of TDP ratings. We are guessing this was done in order to offer the widest range of solutions in a price bracket large OEMs like Dell, HP and Lenovo will likely buy into.

Finally, the new GPU lineup is rounded out by the diminutive GT 415M. This is still considered a “performance-oriented” part due to its respectable 48 CUDA cores but it is still one of the lowest rungs of the mobile GeForce ladder. The only two cards which are lower than it are the 310M and the 305M.
 
Last edited:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
400M-Series TDP: An Explanation & Our Findings

400M-Series TDP: An Explanation & Our Findings


Before we go any further here, let’s state one thing: NVIDIA doesn’t discuss power consumption or TDP of their mobile parts with the press. Or so we’ve been told. Nonetheless, as you can see in the charts below we have dug up many of the TDP numbers NVIDIA gives to system integrators. There is however an explanation that needs to go along with them in order for you to better understand how these numbers are calculated.

To begin with, TDP does NOT equate power consumption so please put that thought to rest once and for all. TDP (thermal design power) is a number given in watts which acts as a guide that allows system integrators to properly design cooling solutions for a given product. Basically, it tells SIs how much heat will be produced and as such, they can configure a chassis and heatsink combination which will effectively keep thermals within acceptable boundaries.

400M-6.jpg

GTX 480M GPU + MXM Module

When it comes to mobile parts, things get even more complicated. In many cases, higher-end graphics cards are integrated into laptops using Mobile PCI Express Modules (MXM for short) which means they can be changed out due to their use of a standardized installation format. These modules act much like the add-in cards we are used to seeing in the desktop market and incorporate the memory and power distribution for the GPU. On the other hand, some vendors integrate the GPU directly onto their motherboard PCB which eliminates the need for a stand-alone module.

The issue with TDP and even power consumption figures in the mobile GPU market is that some include the MXM module in their calculations while others only indicate the power consumption / TDP of the GPU core itself. Naturally, these two numbers will be at odds with one another since the addition of the MXM module will increase heat production as well as power needs. As such, the numbers you see below are the TDP of ONLY THE GPU CORE. Typically, the addition of an MXM module can add anywhere from 10W to 30W more heat to a mobile GPU.

400M-9.jpg

As you can see, the TDPs of the NVIDIA cards are relatively similar to those seen on the ATI side of the fence except when you get into the ultra high end segment where the GTX 480M is the only contender. Remember, the GTX 460M / GTX 470M are meant to compete directly with the HD 5870 so things look quite even here. However, these numbers do not take the MXM modules into account and we already know that the 2GB module the GTX 480 uses pushes it’s TDP close to the 100W mark.

All in all, it looks like NVIDIA has managed to keep the heat production of their Fermi-based mobile parts well within reason if you look below the GTX 480M. How these parts stand up to the competition in performance is another matter altogether though.
 
Last edited:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
Optimus: An Ace up NVIDIA’s Sleeve

Optimus: An Ace up NVIDIA’s Sleeve


6.JPG

There has been quite a bit of talk over the last few months about NVIDIA’s latest technology for notebooks called Optimus. While we didn’t have the opportunity to give you the complete picture when it was first released, we are nonetheless excited about the possibilities it brings to the table for notebook and netbook users alike. It is also a technology that NVIDIA will be bringing to all of their upcoming notebook GPUs.

No matter what kind of mobile user you are, battery life is most likely one of your primary concerns. In the past, there always seemed to be a tradeoff between battery life and graphics performance but there were initiatives to strike a balance. We have seen some technologies which switch between discrete and IGP graphics yet they were anything but seamless; often requiring a system reboot to function properly. Meanwhile, NVIDIA’s Optimus promises to do exactly what previous lacked: a truly seamless transition between the power savings of an integrated graphics processor and the power of a dedicated GPU.

7.JPG

In order to do this on the Intel’s latest platforms, the IGP acts as the display controller so in effect it handles all of the output to the notebook’s monitor. When the dedicated GPU isn’t needed, it shuts off and allows the more efficient on-chip IGP to take over the tasks associated with displaying images on the screen.

8.JPG

Once Optimus detects additional graphics performance is needed, it switches on the GPU. Meanwhile, the IGP still acts as the display controller to the system’s screen but if output is done through an external source via HDMI, it will be the discrete graphics chip that directly outputs the video signal. This is all supposed to happen seamlessly and without the user knowing it is happening behind the scenes. However, it can also be manually switched on and off if you want to fine tune control a bit more.

This dynamic load balancing between the GPU and IGP results in some downright amazing performance to be eked out of a notebook while maintaining extremely good battery life. In their own internal testing, NVIDIA claims they have seen near-equal performance between a notepad without a dedicated GPU and one equipped with a dedicated GPU and Optimus switching capabilities.

Below is our full video review of an Optimus setup. We encourage you to check it out since the features this technology provides are eye opening to say the least.

<object width="640" height="385"><param name="movie" value="http://www.youtube.com/v/frC2SLldzZc&hl=en_US&fs=1&hd=1"></param><param name="allowFullScreen" value="true"></param><param name="allowscriptaccess" value="always"></param><embed src="http://www.youtube.com/v/frC2SLldzZc&hl=en_US&fs=1&hd=1" type="application/x-shockwave-flash" allowscriptaccess="always" allowfullscreen="true" width="640" height="385"></embed></object>​
 

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
Feature Packed GPUs to Go

Feature Packed GPUs to Go


We have already talked about Optimus but there are plenty of other features NVIDIA has packed into their new mobile cards. Many of these we have already discussed at length but in this section we will go through a few of them just to offer up a bit of a refresher.


CUDA, CUDA and more CUDA

400M-4.jpg

CUDA may have been treated as a bit of an outcast in enthusiast circles due to it being excessively marketed when it was released but it seems like most people have begun to see exactly why NVIDIA was so caught up in this technology. Recently, more and more applications that can be accelerated on the GPU have become available and the GeForce product family is reaping the rewards.

One of its most important aspects is the ability to do certain tasks faster than modern CPUs ever could. This should be of particular interest to laptop users since finishing tasks quicker increases productivity and will lower overall battery consumption as well. Everything from photo touch up to HD video editing with Cyberlink’s PowerDirector software can be accelerated by the onboard GPU.

Naturally, game-enhancing features like PhysX also fall under this umbrella but of more interest to laptop buyers will likely be the movement to add GPU acceleration to web browsers. We have already seen the profound effect Flash 10.1’s GPU acceleration had on streaming HD video performance and the addition of WebGL, Direct2D and HTML5 Video will only add to the amazing things GPUs can accomplish in online content.


3D Vision & Blu Ray 3D

400M-3.jpg

We have covered desktop 3D Vision extensively in the past but not many are aware that 3D has been quite successfully ported into the notebook space as well. 3D is everywhere these days with both Hollywood and TV manufacturers cramming content down our throats. However, very few would argue against the ability to game with 3D Vision and watch 3D movies from a laptop and that is exactly what NVIDIA has done.

When gaming in 3D, the performance boost the Fermi-based GPUs have over the previous generation makes displaying what amounts to two full images on the screen much easier. This means a more fluid gameplay experience when using 3D Vision.

400M-1.jpg

NVIDIA has also built into all of their 400-series mobile GPUs support for the 3D Blu-ray format which is an absolute necessity for watching movies in 3D.

400M-2.jpg

3D Vision Play is the final piece of the 3D notebook puzzle and it comes standard on all 3D Vision-equipped notebooks. This piece of software allows the NVIDIA GPU to sync with 3D capable HDTVs via the notebook’s HDMI 1.4 output. As a result, the standard 3D Vision glasses can be made to work with a TV set that would otherwise be incompatible and the laptop can be used as a stand-along 3D Blu-ray player.


Lossless Audio Playback

HD5450-20.jpg
HD5450-21.jpg

All of the mobile GPUs under the GTX 470M also support full bitstreaming of HD audio over HDMI. This means every bit of signal processing is done on the GPU itself without the need for external decoding. This is a huge step forward for those of you who want true high definition audio to go along with a 3D experience.
 
Last edited:

SKYMTL

HardwareCanuck Review Editor
Staff member
Joined
Feb 26, 2007
Messages
12,840
Location
Montreal
Initial Thoughts

Initial Thoughts


There really isn’t all that much to say about NVIDIA’s new mobile 400-series GPUs since we have yet to properly test them against the competition. At this point all we have to go by is NVIDIA’s statements and a whole lot of marketing mumbo jumbo but if promises can be delivered in the retail products, things could become very interesting indeed.

After sifting through more information from OEMs and small system integrators than I care to remember, a few things have become abundantly clear. Both ATI and NVIDIA cut down the specifications of their desktop cards in order to make them more palatable for use in notebooks so don’t expect desktop-like performance from any notebook GPU. The main difference between the 400-series and the HD 5000-series is just how MUCH was cut; particularly from the high-end cards. The GTX 480M holds some semblance of the desktop card but the HD 5870 is a shadow of its former self in mobile form. This circumcision of the high-end ATI card’s performance means NVIDIA has been able to line up their GTX 460M as its equal in performance and integration price. At this point we are a bit less clear as to where the rest of the 400M lineup stand in relation to their competition but an arbitrary “40% faster than the 300M-series” was mentioned in the press material.

We believe that NVIDIA has incorporated enough value-added features into their cards of late to make them serious contenders at every price point. CUDA and Optimus in particular stand out as shining examples of NVIDIA’s innovations having real-world benefits. Meanwhile, Blu-ray 3D and 3D Vision are just two more bits of icing on an already appealing cake.

400M-5.jpg

We do however have some words of caution. From our conversations with SIs and OEMs, it seems like NVIDIA may lead the laptop GPU performance battle up until the next generation of ATI products are released sometime later this year. What NVIDIA needs is to stay price-competitive regardless of how well their GPUs actually perform since the efficiency flag seems to be still firmly planted on ATI’s front lawn. Speaking of efficiency, we believe NVIDIA’s refusal to discuss TDP is utterly counter-intuitive especially when these figures are readily available for the competition’s products. It’s as if they think sweeping these types of discussions under the rug will make people forget that ATI currently has an extremely efficient architecture. Nonetheless, it looks like the TDP of these new NVIDIA cards is quite close to that of the competition which makes the omission of the numbers all the more mind boggling.

It is great to see NVIDIA attacking the mobile computing market with all guns blazing. How these new cards line up with ATI’s solutions is anyone’s guess at this point but in the coming weeks, we intend to find out. If anything, NVIDIA looks to have set in motion an extremely competitive top to bottom DX11 lineup which will definitely have a positive affect on the laptop market.
 
Last edited:
Status
Not open for further replies.

Latest posts

Top