64-bit beta Chrome Web browser arrives for Windows

64-bit beta Chrome Web browser arrives for Windows

Summary: Although Google promises that this new 64-bit browser will be faster and more secure than its current 32-bit version, when it comes to speed, it's not there yet.

SHARE:
10

On June 3, Google announced that it had released a 64-bit version of its Chrome Web browser for Windows 7 and 8. A 64-bit Chrome Web browser has long been available on Linux.

64-bit-Chrome
Google's 64-bit Chrome for Windows 7 and 8 is still very much a beta.

Google promises that it will give users a faster and more secure browsing experience. That may happen one day, but it's not there yet when it comes to speed. 

The program is currently only available in the Canary (read alpha) and dev (read beta) channels. You must deliberately choose to run these. Even if you're running 64-bit Windows today, you will not automatically be "upgraded" to the 64-bit Chrome.

To see how fast it was, I first ran benchmarks on the latest stable release, Chrome 35. I used Sunspider 1.02Peacekeeper, and Octane 2.0, Google's own Web browser benchmark, as well as Mozilla's Kraken 1.1 benchmark. I then installed and ran the 64-bit developer version, Chrome 37.

I ran all these tests on a Gateway DX4710 running Windows 7 SP1. This PC is powered by a 2.5-GHz Intel Core 2 Quad processor and has 6GB of RAM and an Intel GMA (Graphics Media Accelerator) 3100 for graphics. It's hooked to the Internet via a Netgear Gigabit Ethernet switch, which, in turn, is hooked up to a 100Mbps (Megabit per second) cable Internet connection.

To make a long story short, the 64-bit version ran 9 percent slower than its older 32-bit brother on Sunspider; 12 percent slower on Peacemaker; and not quite 2 percent slower on Octane. Oddly enough, the only test the 64-bit edition was quicker on was the Mozilla test where it ran 14 percent faster. In short, if you're looking for a big speed boost today, do not look to the 64-bit version of Chrome.

After the program has matured into a stable release, and the bug-checking code has been removed, it should be a different story. For now, if speed is what you want from a Windows Web browser, stick with the 32-bit Chrome.

Looking ahead, Google promises:

  • Speed: 64-bit allows us to take advantage of the latest processor and compiler optimizations, a more modern instruction set, and a calling convention that allows more function parameters to be passed quickly by registers. As a result, speed is improved, especially in graphics and multimedia content, where we see an average of 25 percent improvement in performance.
  • Security: With Chrome able to take advantage of the latest OS features such as High Entropy ASLR on Windows 8, security is improved on 64-bit platforms as well. Those extra bits also help us better defend against exploitation techniques such as JIT spraying, and improve the effectiveness of our existing security defense features like heap partitioning.
  • Stability: Finally, we’ve observed a marked increase in stability for 64-bit Chrome over 32-bit Chrome. In particular, crash rates for the renderer process (i.e., web content process) are almost half that of 32-bit Chrome.

Again, Google's claims may well be true someday; it's just not true in this first beta release. Only those who like living on the web's cutting edge should use 64-bit Chrome for Windows for now.

Related Stories:

Topics: Networking, Browser, Google

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

10 comments
Log in or register to join the discussion
  • This article

    Was surprisingly neutral in tone. I don't know WHAT I was expecting, but still, it surprised me.
    luke mayson
    • its probably because you don't like google so the criticism seemed neutral

      If he said "windows isn't quite there yet" you'd percieve it as a biased "bashing" of windows
      drwong
      • Except this is balanced

        They made an incorrect claim and he called them out on it. It isn't like he bashed it or anything.
        Michael Alan Goff
  • He has a Windows machine!

    OMG, he has a Windows machine!!!! ....and he obviously using it more than his Linux machines!
    adacosta38
    • He has a Windows machine!

      Where do you get the "more" from?
      Media Whore
  • This does not bode well

    as in the past, with other browsers that have gone from beta to gold, changes to achieve stability and rid bugs make them slower, not faster.
    chrome_slinky@...
  • Does everyone know what Alpha / Beta means?

    "Google's claims may well be true someday; it's just not true in this first beta release."

    Which is why it isn't part of the public, stable release.

    "Google promises that it will give users a faster and more secure browsing experience." - yes it will - that's 'will' as in "in the future"

    From the the author's point of view, there was little point in doing speed bench tests (except to prove it wasn't a 'final' release, which he already knew). From a neutral POV, it does provide a baseline for later comparisons.

    There seems to be significant dumbing down at ZDnet; once upon a time, I wouldn't have had to explain all this.
    Heenan73
    • You're missing the point

      Google is claiming that "Chrome 64" is faster *now*. Thus, SJVN tested their claim and found it wanting.
      Media Whore
  • Re-test needed.

    If you plan on comparing performance, why would you not compare the 32-bit version of chrome 37 with the same version in it's 64-bit flavour? It's far more likely that performance optimisations simply have not been made yet as it's so early in its development process.

    As Chrome 37 gets closer to release as stable, many/most of any performance regressions will be sorted, but it's not very useful to compare chrome 35 32-bit with the 64-bit dev of Chrome 37 without knowing what performance differences are actually the cause of changes between 35 and 37 and not the changes between 32 and 64bit.
    Alan Burns
  • crash rates

    "In particular, crash rates for the renderer process (i.e., web content process) are almost half that of 32-bit Chrome."

    I have to chuckle at this. of course the crash rate for all versions should be zero. Recognizing that software can never be perfekt (I should know, I'm retired from that biz), it still seems to me that if they are aware of a systematic weakness in a certain area, that the code could/should be better protected, so if some anomalous situation arises, it could gracefully handle it vs. crashing and burning. I mean, isn't that the whole idea with separate processes for each tab so the entire browser doesn't go down with the leaky lifeboat?

    It seems to come in waves, and I don't know if it is a function of the exact update version, or the pages I'm visiting, but I have seen a disturbing number of crashes. That said, appears to be a bit better lately (all 32-bit versions). Maybe all those bug reports I've automatically generated ;-).
    ProfQuill