The Fastest Mobile GPU in the World: NVIDIA's GeForce GTX 480M

We'll cut to the chase: NVIDIA's GeForce GTX 480M does indeed reclaim the crown of fastest mobile GPU from ATI's strangely anemic Mobility Radeon HD 5870. And it ought to: after all, in a break with tradition the GeForce GTX 480M is actually properly named. The GPU core of the GeForce GTX 480M is indeed a Fermi GF100 lovingly crammed into a notebook form factor. NVIDIA cut that same 3.1 billion transistor you've all come to know and love down to a still-egregious 100-watt TDP, the highest of any mobile graphics hardware to date.

While the GeForce GTX 285M was just another rehash of the G92 and ATI's Mobility Radeon HD 5870 is a mobile version of their desktop Radeon HD 5770, NVIDIA's GTX 480M uses the same cut-down—but still Fermi—core found in desktop GeForce GTX 465 cards. That means 352 of NVIDIA's "CUDA cores" and a 256-bit memory bus connected to GDDR5 memory. The difference is that while the GTX 465 only gets 1GB of GDDR5, the GTX 480M gets a full 2GB in our review notebook. Clock speeds aren't as comparable, though, with the 480M's clock speed down from the GTX 465's 607 MHz to just 425 MHz. The shader clocks get cut down, too, dropping from 1.2 GHz to 800 MHz. Probably the most alarming drop is the GDDR5: running at 3.2 GHz on the desktop card, the 480M has its effective speed cut to just 2.4 GHz, the lowest speed we've ever seen on GDDR5 and actually a slower effective clock speed than the GDDR3 on the desktop GeForce GTX 285!

It's not unreasonable to expect cuts had to be made to fit Fermi into a notebook form factor, but given how underwhelming the performance of the desktop GTX 465 that the GTX 480M borrows its hardware from is, one can begin to become genuinely concerned about NVIDIA cutting too deep to get the GTX 480M to fit into its 100-watt TDP. Both ATI's Radeon HD 5830 and NVIDIA's GeForce GTX 465 have proven that it's possible to hit a wall of diminishing returns when harvesting high-end GPU die; both of these cards are barely faster than their substantially smaller predecessors.

Still, it's impressive Fermi made it into a notebook form factor at all. 352 of NVIDIA's CUDA cores are nothing to sneeze at, and a 256-bit memory interface connected to 2.4 GHz GDDR5 still produces a healthy 76.8GB/sec of bandwidth. The GeForce GTX 480M also brings to the table full DirectX 11 support along with all of NVIDIA's usual trimmings: CUDA, PhysX, and 3D Vision. With Eyefinity largely out of the equation in notebooks, the Mobility Radeon HD 5870 can certianly begin to feel feature light compared to NVIDIA's hardware.

So that's all well and good, but how does the GeForce GTX 480M actually fare in practice?

Introducing the Fermi-in-Your-Backpack Synthetic Benchmarks
Comments Locked


View All Comments

  • james.jwb - Thursday, July 8, 2010 - link

    Forgot that I use magnification for this site. It's definitely the main cause of the huge performance hit, ouch! (dual-core, pretty fast machine really).

    I think it would be a lot easier if the space now used for the carousel became something static along the lines of Engadget's chunk for "top stories". It's nice to have something there to point out important reviews/news -- I wouldn't want to see the idea completely gone, it's just a carousel is so December 2009 :-)
  • Spoelie - Thursday, July 8, 2010 - link

    While it seems generally true that power keeps increasing from generation to generation (3870, 4870, 5870), wasn't the big drop from the HD2900 series conveniently left out to make that statement stick?

    It's not really that power always increases, there's a ceiling which was reached a few generations ago and the only thing you can say is that the latest generations are generally closer to that ceiling than most of the ones before it. What the desktop GTX480 pulls is about the most what we will ever see in a desktop barring some serious cooling/housing/power redesigns.
  • bennyg - Thursday, July 8, 2010 - link

    2900 was the P4 of the gfx card world regardings power/performance. It was only released because ATi had to have something, anything, in the marketplace. If ATi had as much cash in the bank as did Intel, they would have cancelled the 2900 like Intel did Larrabee.

    Thankfully the 2900 went on from its prematurity to underpin radeons 3, 4 and 5. Whereas Prescott was just brute force attempting to beat thermodynamics. Ask Tejas what won :)
  • JarredWalton - Thursday, July 8, 2010 - link

    That's why I said "generally trending up". When the HD 2900 came out, I'm pretty sure most people had never even considered such a thing as a 1200W PSU. My system from that era has a very large for its time 700W PSU for example. The point of the paragraph is that while desktops have a lot of room for power expansion, there's a pretty set wall on notebooks right now. Not that I really want a 350W power brick.... :)
  • 7Enigma - Thursday, July 8, 2010 - link

    Thank you for the article as many of us (from an interest standpoint and not necessarily from a buyer's standpoint) were waiting for the 480M in the wild.

    My major complaint with the article is that this is essentially a GPU review. Sure it's in a laptop since this is a notebook, but the only thing discussed here was the difference between GPU's.

    With that being the case why is there no POWER CONSUMPTION numbers when gaming? It's been stated for almost every AVA laptop that these are glorified portable desktop computers with batteries that are essentially used only for moving from one outlet to the next.

    I think the biggest potential pitfall for the new 480M is to see with performance only marginally better than the 5870 (disgusts me to even write that name due to the neutered design) is to see how much more power it is drawing from the wall during these gaming scenarios.

    Going along with power usage would be fan noise, of which I see nothing mentioned in the review. Having that much more juice needed under load should surely make the fan noise increased compared to the 5870....right?

    These are two very quick measurements that could be done to beef up the substance of an otherwise good review.
  • 7Enigma - Friday, July 9, 2010 - link

    Really no one else agrees? Guess it's just me then.....
  • JarredWalton - Friday, July 9, 2010 - link

    We're working to get Dustin a power meter. Noise testing requires a bit more hardware so probably not going to have that for the time being unfortunately. I brought this up with Anand, though, and when he gets his meter Dustin can respond (and/or update the article text).
  • 7Enigma - Monday, July 12, 2010 - link

    Thanks Jarred!

    For all the other laptop types I don't think it matters but for these glorified UPS-systems it would be an important factor when purchasing.

    Thanks again for taking the time to respond.
  • therealnickdanger - Thursday, July 8, 2010 - link

    "Presently the 480M isn't supported in CS5; in fact the only NVIDIA hardware supported by the Mercury Playback Engine are the GeForce GTX 285 and several of NVIDIA's expensive workstation-class cards."

    I did the following with my 1GB 9800GT and it's an incredible boost. Multiple HD streams with effects without pausing.

    I figured out how to activate CUDA acceleration without a GTX 285 or Quadro... I'm pretty sure it should work with other 200 GPUs. Note that i'm using 2 monitors and there's a extra tweak to play with CUDA seamlessly with 2 monitors.
    Here are the steps:
    Step 1. Go to the Premiere CS5 installation folder.
    Step 2. Find the file "GPUSniffer.exe" and run it in a command prompt (cmd.exe). You should see something like that:
    Device: 00000000001D4208 has video RAM(MB): 896
    Device: 00000000001D4208 has video RAM(MB): 896
    Vendor string: NVIDIA Corporation
    Renderer string: GeForce GTX 295/PCI/SSE2
    Version string: 3.0.0
    OpenGL version as determined by Extensionator...
    OpenGL Version 2.0
    Supports shaders!
    Supports BGRA -> BGRA Shader
    Supports VUYA Shader -> BGRA
    Supports UYVY/YUYV ->BGRA Shader
    Supports YUV 4:2:0 -> BGRA Shader
    Testing for CUDA support...
    Found 2 devices supporting CUDA.
    CUDA Device # 0 properties -
    CUDA device details:
    Name: GeForce GTX 295 Compute capability: 1.3
    Total Video Memory: 877MB
    CUDA Device # 1 properties -
    CUDA device details:
    Name: GeForce GTX 295 Compute capability: 1.3
    Total Video Memory: 877MB
    CUDA Device # 0 not choosen because it did not match the named list of cards
    Completed shader test!
    Internal return value: 7
    If you look at the last line it says the CUDA device is not chosen because it's not in the named list of card. That's fine. Let's add it.

    Step 3. Find the file: "cuda_supported_cards.txt" and edit it and add your card (take the name from the line: CUDA device details: Name: GeForce GTX 295 Compute capability: 1.3
    So in my case the name to add is: GeForce GTX 295

    Step 4. Save that file and we're almost ready.

    Step 5. Go to your Nvidia Drivercontrol panel (im using the latest 197.45) under "Manage 3D Settings", Click "Add" and browse to your Premiere CS5 install directory and select the executable file: "Adobe Premiere Pro.exe"

    Step 6. In the field "multi-display/mixed-GPU acceleration" switch from "multiple display performance mode" to "compatibilty performance mode"

    Step 7. That's it. Boot Premiere and go to your project setting / general and activate CUDA
  • therealnickdanger - Thursday, July 8, 2010 - link

    Sorry, I should have said for ANY CUDA card with 786MB RAM or more. It's quite remarkable.

Log in

Don't have an account? Sign up now