6 Comments

Summary:

I was thrilled when I saw a tweet about the release of Parallels 6. Earlier I wrote a lengthy evaluation of gaming on virtual machines. I thought I’d run the same games through Parallels 6 to see how its claims work out in the real world.

parallels_game_time_thumb

I was thrilled when I first saw a tweet about the release of Parallels Desktop 6, followed closely by an e-mail from Parallels saying, “Upgrade NOW for the low, low price of $49.95.” In the press release it promised two things that immediately intrigued me:

  • Launch Time: Run Windows applications 41 percent faster than Parallels Desktop 5
  • Enhanced 3D Graphics: Enjoy more life-like visual action and play a wider range of modern games with 40 percent improvement over Parallels Desktop 5

Earlier this year, I wrote a lengthy evaluation of gaming on virtual machines. I thought I’d run the same games through Parallels 6 to see how its claims work out in the real world. Before we get going, though, one quick note: upgrading to this version will force your copy of Windows to re-authenticate, so, if like me you’re near your limit, be warned.

The Test Bed

The fine print. These tests were run on a MacBook Pro 13″ with 4GB of ram, 500GB hard drive and an NVIDIA 9400M graphics chip. The release version of Parallels Desktop 6 was used — at least the version that was available on September 9 — with Windows 7 Professional (the virtual machine was accessing the boot camp parition) with all the latest service patches as my OS of choice. The games tested were Lord of the Rings Online, Dungeons and Dragons Online, and EverQuest 2. These were scientifically chosen from a pool of games that, frankly, are the ones I play a lot. Also, MMOs tend to be the most demanding of games so they make great test subjects. Also, note: Lord of the Rings Online has gone free-to-play, and this or Boot Camp are the only ways to run it currently on a Mac.

Claim One: Improved Launch Times

For the most part, the claims held true. Without breaking out a calculator, I did find the load times for games to be better on Parallels 6 over Parallels 5. Interestingly, EverQuest 2 was slightly longer on the load times with version 6. The OS itself took exactly the same time to load on both versions. The oddity I mentioned in the earlier article, where it took Parallels longer to launch on a clean boot of OS X, still holds true.

Claim Two: Increased Performance

After seeing increased load performance, I expected to also see much better FPS. Now, the FPS tests aren’t scientific; none of the games have a true benchmark test. If I’d been thinking straight, I’d have run Speedmark tests on both versions. However, I’ve never really trusted the benchmark tests over actual in-game observations.

I did not see the increased performance Parallels was claiming. The tests were pretty much a wash, with both versions testing within any margin of error of each other. They may appear slightly better or slightly worse than the previous version, but statistically it’s a wash.

The Intangibles

Version 6 has support for 64-bit and 5.1 surround sound. Frankly, I couldn’t tell a difference with the 64-bit support and I don’t have 5.1 surround on my MacBook.

Conclusion

Before version 5, upgrading Parallels quickly became a regretted decision. It felt like things were worse, not better, and I’d have a host of performance-related issues. That’s changed, and after an afternoon with Parallels 5 I don’t regret upgrading.

That said, I can’t really tell a difference. It does feel slightly more responsive, but I can’t measure how. I really wish there was quantitative way to say “It’s snappier!”

Do I recommend upgrading if you’re a gamer? Yeah, I do. While “It didn’t make things worse” is hardly an enthusiastic recommendation, I’ll be the first to admit that testing three games by two different developers is an incredibly small sample set. That said, since upgrading is a one-way street (I’ve never had much luck backing down a version), I can’t recommend checking the Windows Guest forums at Parallels’s official site before upgrading. There is no shame in letting people like myself be your canaries.

You’re subscribed! If you like, you can update your settings

  1. I was on the fence about this because it was not clear what the advantage was, but then MacUpdate started running a $39 special where you get a free upgrade to 6 on release day, so I figured I’d not find it cheaper and might possibly find it better.

  2. Did you mean version 6 in the conclusion?

  3. Maybe the improved performance in games would be more obvious if you had used a machine with a dedicated GPU rather than a 9400M.

  4. And where comparison with Boot Camp or VMware for this games.
    The article whould be a lot interesting if you tried harder.

  5. Any 3d tests with the 9400M aren’t going to be that revealing. You can’t squeeze blood from a stone.

    Maybe the 9600M

  6. Your FPS tests are BEYOND USELESS. You CANNOT test FPS by starting a game and moving around inside of it. FPS will vary greatly by such simple factors as what DIRECTION you are looking in. This is because 3D rendering and the resulting FPS isn’t some kind of static process. It depends COMPLETELY on what you’re looking at, how many characters are on screen, and so on, or to put it more technically: More geometry on screen = lower FPS.

    You should do something like the Crysis benchmark or 3DMark and you WILL see the massive improvement.

    ArsTechnica did the Crysis benchmark and it jumped from 22.0 to 36.2 FPS. A hefty upgrade. An upgrade to your blog post after you’ve used actual, valid, unbiased testing is advisable. ;-)

Comments have been disabled for this post