ASUS UL50Vf Review Discussion

The new ASUS UL50Vf is an update to the popular 15-inch UL50Vt and offers the all-new Nvidia Optimus switchable graphics. Why are we reviewing this modest update? Because Optimus is honestly the first switchable graphics solution that works like it should.

Read the full content of this Article: ASUS UL50Vf Review

Related Articles:

• ASUS UL80Vt-A1 Review
• ASUS UL30 Review

 

Long video on Nvidia Optimus technology.

 

very nice review!!!Also very quick. This laptop and technology has just been announced!

 

We have to thank Nvidia for helping us get our hands on this review unit in advance of the launch date. We love those guys ... even if we sometimes recommend ATI hardware.

As I said in the review, it's not the most impressive ASUS notebook we've seen in our lab, but we are very impressed with the Nvidia Optimus technology.

It makes switchable graphics a no-brainer for average users and still gives experienced users the control they need to override the automatic settings.

This makes a lot more sense than always running on the IGP or always running on the GPU unless you remember to switch.

 

What else has Asus revealed to you ... that you're not telling us

Bronsky

 

You know the old saying:

"Those who say don't really know, and those who really know can't say ... because of non-disclosure agreements they've been forced to sign in order to find out in advance."

 

So... you review a laptop with Optimus technology, and don't even get around to explaining what Optimus is and how it works until page 2?

On top of that, the explanations are sparse and there is no objective testing of how well the Optimus system works. A benchmark (or two, or three, or five) and some battery life figures comparing the system with Optimus on to discrete GPU only and IGP only would have been nice.

You make a big deal about Optimus (and I agree), but then never get around to explaining or showing (even more important) why it is so good. A notebook review showcasing new and innovative technology should not be given the same treatment as copy and paste reviews for commodity products. I hope that NBR revisits and expands the existing review, because as it is, it is inadequate.

 

We like to use the forums to go into that level of detail because, let's be perfectly honest ... the overwhelming majority of average consumers who will read this review don't "really" care about the technical details of how Nvidia Optimus works. They just want to know that it works.

I guarantee that the vast majority of people driving hybrid cars don't know the technical/mechanical details of how hybrid gasoline/electric engines work ... but that doesn't stop those people from buying hybrid cars.

For example, Nvidia did some research among people who currently own notebooks with switchable graphics and Nvidia claims that only about ONE PERCENT of current switchable graphics owners switch back and forth between IGP and GPU. Most consumers either leave their notebooks set to the factory default or they switch it one time after buying the notebook (either to the IGP or the GPU) and leave it that way.

For those of our forum members who want the details of how Optimus works, it's pretty simple: Nvidia eliminated the use of multiplexers (muxes).

Current switchable graphics requires the use of multiple muxes on the motherboard to handle the mutliple outputs from the Intel IGP and the GPU (either Nvidia or ATI) and channel it to a single output for the display(s) (one mux for the internal display and one mux for each external display port). This means motherboards have to have multiple muxes which take up space and consume battery power.

This is why current switchable graphics require either a system reboot or a soft restart of the display driver ... because the muxes have to switch between the IGP and the GPU for display output.

Nvidia's solution with Optimus was to remove the muxes and make the IGP the single output source for the display(s). The IGP is always running (just like with current switchable graphics) but the discrete GPU sends its output through the IGP and the IGP is "tricked" into thinking it's doing the work even when the GPU is actually rendering what goes on the screen. When the GPU isn't needed for an application it is simply turned off and the IGP handles everything. When the Optimus driver detects a software call for directX or any similar GPU-required call it turns on the GPU and the GPU does the work and sends everything to the display through the IGP.

The switching is therefore handled entirely by software (activating or deactivating the GPU based on whether the software/application you're running needs the GPU or whether you manually set an application to require the GPU/IGP by right clicking on the application and setting a default graphics option).

If you'd like additional technical details I'm happy to provide them here in the forums, but I hope you understand why this wasn't included in the main text of the review ... we want readers to actually read our reviews and not get bored and skip them.

Let me know how much technical detail you want and I'll be happy to provide it here.

 

Great review! If only the build quality wasn't as questionable. (or the price was a bit lower to reflect that....)

Hopefully more laptops will use this switchable graphics technology.

 

Very nice review ... Asus build quality is not stupendous, it is true. But it is acceptable imo.

I am also all for switchable graphics technology!

 

I think a small technical summary section of Optimus and how it works should be in the review part since I don't always get to read everything in the forums

 

In addition to my previous post explaining that Nvidia Optimus works by eliminating multiplexers (muxes) from the motherboard and uses the IGP as the single display controller (tricking the IGP into thinking it's doing the work even when the Nvidia GPU is actually handling the rendering), here are some additional technical details to keep our technical readers happy.

Optimus uses standard industry protocols and Microsoft APIs to communicate with the Intel IGP driver and uses the PCI-Express bus to transfer the Nvidia GPU's output through the Intel IGP. In short, Optimus is a combination of hardware and software solutions.

The software element in Optimus is a combination of the "Optimus Routing Layer" and the "Optimus Profiles." The Optimus Routing Layer is part of the Nvidia GPU driver and includes a kernel-level library for recognizing and managing specific applications associated with different graphics hardware (IGP or GPU). Specifically, the Optimus Routing Layer determines when the GPU is needed based on three call types:

1. DX Calls: Any 3D game or DirectX application will trigger these calls.
2. DirectX Video Acceleration (DXVA) Calls: Video playback will trigger these calls
3. CUDA Calls: CUDA applications will trigger these calls.

The other software element is the "Optimus Profiles" these are simple application profiles (similar to antivirus software updates) that identify whether an application should use the IGP or GPU. Nvidia profiles applications based on a verification/validation process and then updated profiles are hosted on an Nvidia web server and automatically pushed to Optimus users (unless the notebook user has turned off the automatic Optimus updates ... similar to how automatic updates work on antivirus programs).

Another software element to Optimus that isn't specific to Nvidia is Windows 7. Optimus relies on the ability of Windows 7 to allow two independent graphics drivers to be active at the same time. The Intel IGP driver and the Nvidia driver run at the same time. If you look at the Windows Device Manager you'll see two display adapters listed even if the Nvidia GPU is turned off (that isn't the case with previous switchable graphics solutions).

The primary hardware element in Optimus is a new feature called the " Optimus Copy Engine" (OCE). The OCE allows the removal of the multiplexer (mux) by transferring the display surface from the GPU frame buffer over the PCI Express bus to the system memory-based framebuffer used by the Intel IGP.

The OCE is an alternative to traditional Direct Memory Access (DMA) transfers between the GPU framebuffer memory and the system memory used by the IGP. The OCE uses the bidirectional bandwidth of the PCI Express bus to allow simultaneous 3D rendering and copying of display data from the GPU framebuffer to the system memory area used for the IGP framebuffer. The IGP then reads the display data from the framebuffer and sends it to the display interface (built-in display, HDMI, VGA, etc.) as if the IGP had done the work ... even though the GPU actually did the rendering. The asynchronous DMA operation provides the data coherency needed to allow this to happen.

Additional points:

When using current (older) switchable graphics solutions, users were forced to manually enable the GPU before launching any application that needed the GPU. Users also had to ensure that no "blocking applications" were open before manually enabling the GPU (blocking applications include any application that makes a GPU call ... including some basic applications like Solitaire). If a blocking application was open the switch from IGP to GPU would not happen (and this wasn't always obvious to average consumers).

So, what are the differences between traditional switchable graphics and Optimus?

1. Optimus automatically enables the GPU for applications that need the GPU and disables the GPU when you don't need it. Traditional switchable graphics require you to manually switch.
2. Optimus provides "nearly" instant transition time between the IGP and GPU while older switchable graphics solutions require a system reboot or a display driver restart.
3. Optimus provides "flicker-free" switching compared to traditional solutions where the screen turns black while the new driver loads.
4. Optimus doesn't care about "blocking applications" but older switchable solutions don't work if you have a blocking application running.
5. Optimus uses standard drivers. Older switchable graphics require specific drivers to be used and you can't update drivers without potential problems.
6. Optimus doesn't require additional hardware. Older switchable graphics require muxes that take up space and consume battery power.
7. Optimus uses standard APIs and protocols unlike older switchable graphics that use specialized drivers and related software.

 

No, the build quality is normally very good....just not on the UL series.

 

I gotta agree with Jerry on this one, I consider myself a mild computer enthusiast and that explanation bored me, I skipped over the thorough details of how Optimus works.

and the X83VB that I owned. Asus should take notes from HP and their DM3, no one wants glossy, black plastic.

 

Now there's no excuse for all notebooks with a dedicated nVidia GPU to not be switchable, especially those with Arrandales. I think it is perfectly suited for gaming notebooks, e.g., my G50Vt which games nicely but has terrible battery life.

 

Great review and some truly ground-breaking technology from Nvidia which sets the new standard in switchable graphics design.

Im still interested in why there is no option that enables the user to override all the optimus profiles and force the use of only the igp for all apps if he so desires. Im envisaging a scenario here in which a user may need to prolong battery life for the longest possible time by disabling the discrete gpu altogether to conserve as much power as possible and obtain optimal endurance.

 

I understand that argument, but from a real-life usage point of view, I can't imagine a situation where a power user would find Optimus activating the GPU without his/her knowledge.

Optimus uses only the IGP unless you launch an application that needs the GPU ... and you can override that by right-clicking an application and telling Optimus to use only the IGP. In that case, if it's an application that "requires" the GPU it just won't run and if the IGP can handle it then the IGP will handle it.

For example, if I'm on a flight(s) from NYC to Seattle and I'm using a notebook with Optimus and want to get as much battery life as possible then chances are I'm just browsing the web and using Microsoft Office (which would just use the IGP). If I decide to watch a DVD (which I probably shouldn't be doing anyway if I want extended batery life) I can right-click on the movie player application and force it to go to IGP only.

It's also possible that a future Nvidia driver would allow you to "force IGP on all" applications, but that sort of defeats the whole purpose of the Optimus switchable graphics solution that tries to give you the best balance of IGP battery life and GPU performance.

 

Well doesn't the Vaio Z offers the same thing but with an actual switch to either forcing the intel igp, nvidia gpu, and of course auto? Couldn't Asus do a similar thing?

 

The "auto" mode on the VAIO Z is basically a preset to switch the graphics based on battery (IGP) or AC power (GPU) ... unles you override the default settings. It can't detect when an application needs to use the GPU or not, and it still flashes the screen as it switches between the drivers.

I suppose there's nothing to stop ASUS or another manufacturer from using a dedicated button with Optimus, but it again sort of defeats the point of a switchable graphics solution that handles the switching for you so you don't have to worry/think about it.

 

Ive got to agree with coldmack here. For ultimate flexibility, there should be an option to choose between (1).intergrated only, (2) optimus (automatic) or (3) discrete only. These options could surely be easily intergrated at a software level, without having to use a hard switch.

 

You guys don't post the review in the forum anymore, so that you can make more advertising revenue....very clever.

 

Jerry, thanks for your responses to request for technical content. While there are sites that are technical (and I've been reading their articles on Optimus), my main complaint about the article is that it doesn't show consumers how Optimus is any better than IGP, discrete GPUs and switchable graphics.

There are definitely situations where Optimus is activating inappropriately. TR found that the nvidia GPU was activated by applications that were CUDA/DXVA accelerated, even when the app was doing nothing. Examples include flash video not playing, or even just launching Badaboom and keeping it in the background without loading any video files. Also, there are cases where Optimus fails to activate, but that is easily rectified with custom profiles.

Also, is there any noticeable glitch or delay when Optimus switches GPUs?

I'm actually quite impressed with what I've read about Optimus, I just don't think the NBR article does it much credit, as it doesn't make clear to the reader the benefits of hybrid switching.

 

Believe it or not, it is NOT due to ad revenue ... it's actually because of Google.

Our Google page ranking was slipping because Google was seeing the reviews on our main site and the identical reviews posted in the forums as duplicate content.

I'm sure our advertisers would love it if you clicked on those ads, but our editorial staff (myself included) only care about readers being able to find our content on the first page of Google results.

 

As mentioned in the review, there is a one or two second delay in our particular review sample when you launch an application as the Optimus driver figures out if the IGP or GPU should be used and turns on the GPU if needed.

It's unclear how much this delay can be minimized via driver updates from Nvidia. Nvidia's own technical white papers admit there is some delay as the system determines whether the IGP or GPU should be used, but it's a VERY minor delay. It's obviously MUCH faster than traditional switchable graphics where the GPU/IGP driver has to shut down and the other driver has to load, but it isn't as immediate as it would be if the GPU was already on and you launched an application.

 

Pretty neat article. To be fair I skimmed the review except for the Optimus part and the technical details on it here in the forum. Neat stuff in my opinion. But then again I am a computer architecture student.

I agree with Jerry that on-the-fly switching is preferable to a manual switch. Sure it would be nice to have the option of always-IGP, but I know from experience that I'd forget it sometimes. I once almost thought my GPU wasn't working once because I'd turned off DirectX GPU acceleration for a really old game the previous day and had forgotten to turn it back on. I occasionally forget if my wireless switch is on/off as well. And turning off DirectX acceleration in dxdiag will do the job most of the way if you really want to make sure you don't inadvertantly run down the battery.

I'd be interested in reading the whitepapers from nVIDIA, but can't find the link right now. Maybe it'll be easier once all the news stories leave the front pages of Google, but I'd appreciate a link as well - if it's public, of course.

I'm also curious which drivers support this. Clearly some version of nVIDIA's Windows 7 drivers will, and Vista drivers will not as Vista doesn't support multiple graphics drivers being active at the same time, but it's unclear from what I can find if nVIDIA's XP drivers will support this. Technically I'd think it would be possible, as XP does support multiple graphics drivers being active at the same time, but without a test system or information either way, it's difficult to tell. I'd guess that the IGP would be used on a system that without supporting drivers, since the dedicated GPU is going through the IGP, but I'd also be interested in confirmation of that.

I actually expected this to be about some sort of system where the user could switch dedicated graphics, either like the maligned MXM, or like Fujitsu's or ATI's external GPU systems. Considering how little traction those have gained, though, this is probably going to be more relevant.

This sounds like it could be good Apple rumor fodder, too. If it works with Arrandale and not just X4500, then what better way to update to Arrandale and still get nVIDIA graphics performance with integrated battery life? Might take some driver work for OSX, but Cupertino probably wouldn't want Redmond to gain much of an edge with this.

 

@Voldenuit: I just had a chance to look at TR's coverage of Optimus and one thing just doesn't make sense to me. Why would you launch Badaboom, an application that REQUIRES Nvidia's CUDA to launch, if you didn't want the Nvidia GPU on?

Badaboom just errors out if you try to launch it on a notebook without a CUDA-capable Nvidia GPU, so why would you launch that application (even leave it open in the background) if you only wanted to use the Intel IGP and get the most battery life possible?

I'll go ahead and admit it would be nice for Nvidia to add a "disable GPU" button to the Nvidia driver console to keep Optimus from turning on the GPU, but why would you try to launch a GPU-only application if you didn't want the GPU running?

@Apollo13: The Optimus whitepapers (at least the ones I read) are hosted on a secure Nvidia FTP site that requires a password to access. If Nvidia puts them on a publicly accessible website I'll be happy to publish a link here, but I don't want to publish them here if Nvidia hasn't put them on a public website.

That said, the information I've put in this forum thread contains MUCH of the whitepaper data.

As far as I know, there is nothing stopping Nvidia from making Optimus work with Windows XP since, like Windows 7, it allows for multiple display drivers to work at the same time. However, Nvidia said they are only supporting Windows 7 at this time (that could change as soon as Nvidia releases a compatible driver).

The beauty of Optimus is that it works with ANY Intel IGP driver being used by the IGP and you just have to have an Optimus-compatible Nvidia driver for your particular Optimus GPU (GeForce 200M series, 300M series, and future Nvidia GPUs) ... and the motherboard obviously has to be setup to run Optimus (you can't convert a current switchable graphics notebook that uses muxes to an Optimus system with a simple driver update).

As for Apple rumor fodder, Nvidia was quick to say that there is nothing stopping Optimus from working in a Mac OS X environment. Then Nvidia reps were quick to say they know nothing about future Apple products and any questions about Apple products should be directed to Apple.

 

Battery life should have been included in this review. Isn't that the point of having switchable video? If the faster video used as little power as the built-in graphics, there would be no need for the gimmick.

As to why anyone would want lower pixel count in a bigger screen, ask any senior. Or, like you said, games could only be run in low to mid resolution anyway.

 

Battery life is mentioned on the second page of the review (8 hours and 53 minutes based on our testing procedure). That is with the system running in IGP mode only since our standard battery life test doesn't stress graphics.

 

1. Do we know yet what new laptops will have Optimus? Sounds like the Z will NOT and it would require a hardware update to add Opptimus.

2. The newer Intel HD integrated graphics that comes with the new arrendale chips is a reasonably capable IGP (benchamarks 250% better than the previous 4500MHD.). It sounds like Optimus would bypass the IGP in some circumstances where it would work fine. This is where a physical hardware button (or software all IGP) switch would help.

For example, the new IGP should be fine for most Java, video and even some games. It would be nice to be able to easily disable optimus during a flight and then turn it on if you ran into stuttering from the IGP.

 

heh heh heh

Clever nonetheless.

It actually does not bother me as I come to the forums not through the "splash page", per se, but through the forum interface.

 

Answer 1: ASUS was the first to announce new notebooks with Nvidia Optimus (the UL50Vf being the first) but others will certainly follow. Nvidia works with all the major notebook OEMs

Answer 2: Well, you could use profiles/right clicking on apps to force them to use the IGP but the way Optimus works is that it's looking for GPU "calls" from the software that you're running. So if your software asks for DirectX or CUDA Optimus assumes it should be using the GPU and not the IGP. Again, you can override this with profiles and there's nothing to stop a notebook OEM from making an Optimus-based notebook with a button that tells the Nvidia driver to only use the IGP.

Also, related to #2: Nvidia states that Optimus is currently setup to work with Intel "Arrandale" (Core i3, i5, i7) CPUs, Core 2 Duo "Penryn" CPUs and Atom N4xx "Pine Trail" CPUs. So it sounds like the folks at Nvidia are ready for Optimus to work with any Intel IGP being used with Arrandale.

 

That's okay, I figured that might be the case when Google and nvidia.com didn't turn anything up. Thanks for all the information and answering all these questions, though - this has been a really in-depth, useful review especially with the follow-up notes.

I'd +rep you again, but the forum says I need to spread some reputation around a bit, so this'll have to suffice for now.

 

Very, very interesting ... thanks for the indepth report on this!

 

Very nice review - I was actually down at my local Microcenter today (wanted to check out the new HP TouchSmart tm2 - which was very nice, btw), and decided to check out the UL30 they had there. It sounds like the UL30 has similar build quality issues as the UL50 does - I had been very interested in the UL series, but on the UL30 I also noticed a high degree of flex, especially on the keyboard - it was especially striking after trying out the tm2, which had a rock-solid keyboard by comparison.

The Optimus tech does sound very cool - looking forward to seeing it on more laptops.

 

Blargh. Why the heck does it have such a weak build...i would rather have it a bit thicker while knowing that i don't have to worry for Notebook not to be damaged all the time...

Thinner is not better....i hope this won't become a trend....

 

think they will make one in a higher resolution?

 

While I am all for more options in the switchable graphics, just what in the world was on Asus's mind when they decided to put a switchable solution with a GPU of only GT210M's caliber in a 15" laptop? Come on, it's not like this is a 13" like the 2008 Sony Z that had limited options to manage the heat.

Another thing that I am kind of wondering is if (probably more when than if, really) will Intel decide to make its IGP "smarter" and not allow nVidia GPU to flow through it...

 

About the second point, I doubt it will ever happen. On the first point , I agree with you - especially since the latest NV GPUs are pretty power-efficient, and latest Intel integrated GPU are getting closer to the performances of low-end discrete GPUs. It seems rather pointless to have both.

 

hey guys any one has any inside news/scoop at to when is Asus putting in the Core i5 520um or Core i7 620/640um in UL50vf Series...?!?
__________________