liliputing logo
  • Home
  • Products
  • Top stories
  • Reviews
  • Deals
  • Ultrabooks
  • Contact
  • About
 

NVIDIA ION, Broadcom HD video accelerator compared

  • Tweet
  • Email

Credit: Laptop Magazine

Intel Atom powered netbooks typically have more than enough processing power to handle day to day tasks such as surfing the web, editing documents, or watching standard definition YouTube videos. But the integrated graphics solutions that come with these netbooks typically aren’t powerful enough to handle 1080p HD video, high definition Flash video, or hard core video gaming.

Of course, the simplest way to make sure you have a laptop powerful enough to do those things is to either shell out the big bucks for a high end ultraportable or buy a larger laptop that has a netbook-like price, but higher performance graphics and CPU options.

But there’s another approach that some netbook makers have been taking. NVIDIA and Broadcom both offer products that can enable higher performance graphics in one way or another on netbooks with Intel Atom processors. The NVIDIA ION solution replaces the integrated graphics with an NVIDIA graphics processor which adds support for HD video playback as well as the ability to play some video games and perform other tasks that take advantage of the more powerful GPU.

The Broadcom Crystal HD video accelerator is a cheaper option which adds 1080p HD video playback capabilities, but which isn’t a full fledged GPU in its own right. Computers with this card still use Intel’s integrated graphics, but have the ability to tap the coprocessor when it’s time to display HD video.

So how do the two stack up against one another? The best way to test would be to take two machines that are identical in all ways except that one has the Broadcom card while the other has NVIDIA ION. Unfortunately, I’m not aware of any PC maker that offers those two options on the same machine. But the folks at Laptop Magazine did the next best thing and benchmarked a Dell Inspiron Mini 10 with a Broadcom HD card against an HP Mini 311 with NVIDIA ION graphics. For good measure, they threw in a version of the Inspiron Mini 10 without the Broadcom card.

The results, as you can see in the graph above are pretty stunning. While the NVIDIA ION card outperformed the other two options in every test, the Broadcom card enabled 1080p HD video performance on the Dell laptop that would have been impossible without the card. And the difference between 59 frames per second and 48 frames per second is actually pretty insignificant, since your eye won’t notice the difference.

Of course, the Broadcom card doesn’t support video gaming. And the latest version of Adobe Flash Player doesn’t work with the Broadcom HD Video Accelerator yet, which means that HD Flash video isn’t support. But it should be soon.

If you're new here, you may want to subscribe to our RSS feed, follow us on Twitter, or "like" us on Facebook. Thanks for visiting!

Posted on Friday, January 22nd, 2010, 2:01 pm by Brad Linder




  • Kobaljov

    And don't forget the digital display connectors on the Ion. The new dual core D510 Atom boards with the Broadcom chip for the first sight seems to be a good combination for HTPC as of the full passive cooling and full HD playing , but it's lacks the support of resolutions bigger than 720p on the digital connectors (usually the digital connector is missing too). Probably the Intel didn't want to compete with his own lower Core i3 / CULV models.

  • changturkey

    Too bad that most netbooks even with the option to get the Broadcom Card don't have HDMI.

  • Chance_Stevens

    When it comes down to it, there are two different solutions. ION is a full featured solution for gaming, streaming and encoding while the Broadcom card is a $20 add on for people who want to watch online video. I doubt you'd ever see an ION netbook without an HDMI port because it would defeat the purpose of having the graphics.

    I'd really like to see what happens when you take a computer like the Acer 1410 and put it up against a 1201N or Mini 311. The Intel 4500 isn't the 9400M but with a faster processor I'd be interested in seeing the difference.

  • Sam

    If the video has significant 60hz material, you'll notice the difference between hardware that can display at the encoded frame rate and something that can't–image will stutter. This'll probably be most noticable for moderate to slow camera pans. Any smooth motion will show it, but something occupying the full screen makes it really obvious.

    This is why action gamers want graphics cards that can (almost) always render faster than their chosen screen refresh rate, something which renders at an average of 48hz will be horribly stuttery with a 60hz refresh rate.

  • Sam

    Looking at that graph again, I'm wondering if Laptop magazine ran into the limitations of the software they were using. I believe all three clips are actually encoded at 24 fps.

    Then 48 fps from Crystal HD and 60 fps from ION would essentially be decoding as fast as needed by source material, the detected frame rates simply being odd due to the software misdetecting what's going on (differently due to the different hardware architectures). Both solutions are simply being limited by the video playback waiting until the next frame needs to be shown.

    To get a more accurate picture of the video performance, either measure idle time of the decoding hardware (probably not easy, measuring CPU idle may be irrelevant) or use software which decodes frames but doesn't wait to decode the next frame (3D first person shooter gamers know this as not syncing to the display refresh).

  • Sam P

    If the video has significant 60hz material, you'll notice the difference between hardware that can display at the encoded frame rate and something that can't–image will stutter. This'll probably be most noticable for moderate to slow camera pans. Any smooth motion will show it, but something occupying the full screen makes it really obvious.

    This is why action gamers want graphics cards that can (almost) always render faster than their chosen screen refresh rate, something which renders at an average of 48hz will be horribly stuttery with a 60hz refresh rate.

  • Sam P

    Looking at that graph again, I'm wondering if Laptop magazine ran into the limitations of the software they were using. I believe all three clips are actually encoded at 24 fps.

    Then 48 fps from Crystal HD and 60 fps from ION would essentially be decoding as fast as needed by source material, the detected frame rates simply being odd due to the software misdetecting what's going on (differently due to the different hardware architectures). Both solutions are simply being limited by the video playback waiting until the next frame needs to be shown.

    To get a more accurate picture of the video performance, either measure idle time of the decoding hardware (probably not easy, measuring CPU idle may be irrelevant) or use software which decodes frames but doesn't wait to decode the next frame (3D first person shooter gamers know this as not syncing to the display refresh).

  • Pingback: netbook - atom N270

Facebook Twitter Gplus YouTube RSS

Recent Posts

  • Homemade Mondrian PC case shows that mini-PCs can be pretty
  • Cisco pulls the plug on Cius tablet for enterprise
  • HP Pavilion dm1 gets a Brazos 2.0 CPU upgrade… in Malaysia

Latest Products



Popular Discussions

Powered by Disqus

Featured Videos

  • Samsung Galaxy Tab 2 (7.0) video review - Liliputing
    Samsung Galaxy Tab 2 (7.0) video review - Liliputing
  • Kupa X11 tablet with Windows 8 Preview
    Kupa X11 tablet with Windows 8 Preview
mobiputing logo

Latest news from Mobiputing

  • Absinthe 2.0 untethered jailbreak for iOS 5.1.1 now available
  • Unofficially unlock AT&T HTC One X bootloader (with official tools)
  • Android apps can now use in-app billing for subscription content



2007-2012 Liliputing

Advertising | Privacy | TOP