Firefox 4 as Psychedelic as IE9 with Direct2D enabled

IE9 is much faster than Firefox 4 beta at the Psychedelic test on the testdrive site, which demonstrates drawing fast graphics to the Canvas element. That said, a comment to an earlier post prompted me to try enabling Direct2D in Firefox 4.

As you would expect, the difference is dramatic. Here’s the before and after:

image image

How about IE9?

image

Really nothing in it. It’s all about hardware-accelerated graphics.

Direct2D is not enabled by default (currently), so it is not unreasonable for Microsoft to show the slower speed in its published comparison; but worth noting that the issue is easily fixed, presuming Firefox 4 is stable in this mode.

If you want to enable Direct2D in Firefox 4, the how-to is here.

VN:F [1.9.18_1163]
Rate this post
Rating: 9.5/10 (4 votes cast)
Firefox 4 as Psychedelic as IE9 with Direct2D enabled, 9.5 out of 10 based on 4 ratings

Related posts:

  1. Why you should keep UAC enabled on Vista
  2. Programmatically detecting whether UAC is enabled
  3. Browser wars heat up as Firefox 4 arrives
  4. Google Chrome Mac and Linux arrives – may hurt Firefox more than Safari
  5. Phishing part 2: Firefox gets there first

7 comments to Firefox 4 as Psychedelic as IE9 with Direct2D enabled

  • we’re finding that FF 4 is *almost* as fast as IE 9 with acceleration turned on on pretty nearly all of the tests, but IE always has a slight edge – and FF doesn’t get the canvas demos right. Will be interesting to see if FF moves from hoping to committing to ship acceleration and if only accelerating portions of the page rather than the all-in IE approach keep up the near speed parity on real apps vs these less complex demos…

  • Norm

    I was playing with these tests, and I notice the numbers change rather significantly dependent on the window canvas size. I know I’m being picky, but I’d just like to point out the canvas size as shown in the screen shots for your tests are not exactly the same.

  • Destraht

    They could always release 4 without acceleration and then release a 4.1 in 3 months that has it.

  • But what about Linux? Is OpenGL support going to be included for this? If not it could leave us with the biggest disadvantage to date compared to Windows.

    We are already lagging behind with a lack of video acceleration in FLASH because a certain company couldn’t be bothered to work around the issues. Firefox however at least only has to support OpenGL whereas FLASH would have needed a few tricks up its sleeves, as there are about 4 different APIs it would have to support to cover most GPUs.

  • Theta

    @Alex

    No one cares about Linux.

  • Mark

    Have no fear Linux and Mac users.
    Firefox will use opengl layers to provide hardware acceleration on those platforms.
    I just ran Firefox 4b5pre on Linux with layers enabled on a AMD Athlon 64 4000+ dual core processor and I got 1764 revolutions per minute on this test. :)

    In order for opengl layers to work, you must have a video card and drivers that support at least opengl 2.0.
    I was using the Binary Nvidia Linux Drivers.

  • Actually, @Theta, I care about Linux (as, I presume, does Alex), so your glib statement is false.

    To be honest, I don’t care how well IE9 performs because I don’t use Microsoft Windows at all. I personally suspect that IE9 isn’t going to help Microsoft gain any credibility in the browser space. They’re already seen by all the developers I know as hopelessly behind the curve. They continually disappoint due to their internal cognitive dissonance: the world wants cross-platform standards, Microsoft wants total control. They fight among themselves, and that friction allows tiny little but very purposeful open source dev teams eat their lunch. Of course, those paid by Google are also doing quite well and their work is also open source (I’m running Chromium as well. Nice browser). Ultimately, if the developers don’t bother with the MS technologies, why would they bother running Windows? Seems like a nice virtuous circle to me.