Home > computer, hackintosh, macintosh, software, windows > VMware Vs Virtualbox, round 2

VMware Vs Virtualbox, round 2

VMware fusionVirtualboxIn the first part of my testing of VMware Fusion and Innotek Virtualbox I looked at boot time and geekbench scores. Since Geekbench had issues on VMware (and I still haven’t got to the bottom of that) I have moved onto real world tests. I used the Adobe Photoshop CS3 trial and VLC for the testing. I’m fairly certain everyone has heard of Photoshop, and VLC is an open source media player for a wide range of operating systems.

Photoshop tests first, and I ran three tests using three different filters on a common subject. This was a 5,000 by 5,000 pixel RGB image filled with the first of the available patterns. The filters and results are shown below.

Platform
VMware
Virtualbox
Macintosh
Radial Blur 100%
31
31
11*
Charcoal filter 1,5,50
8
8
7*
Water Paper filter 15,60,80
13
10
13*

* – Mac filter used all four cores

* – Mac filter used one core

Some interesting results again. VMware and Virtualbox are well matched in the first two tests, and Virtual box is a bit faster on the last. The strange thing here was the results from running the tests using CS3 natively on the Mac. The first test was way faster than either virtual machine which makes sense as they are using one core and the Radial Blur filter uses all four available cores on the Mac. The Charcoal filter appears to only use one core so was slightly faster on the Mac. The Water Paper filter again only uses one core, but VMware matched the Mac and Virtualbox beat both. I ran the test multiple times to confirm the results and still cannot see how the virtual machine beat the Mac. Is the Windows version of CS3 faster than the Mac version? The only way I can see of getting a genuine comparison will be to install Windows XP natively on the hackintosh and run the tests again.

The VLC test was next, and for this I used VLC’s conversion abilities to produce a file with Divx3 3072Kb/s video and MP3 192Kb/s audio. The source file was a 1280×720 Mpeg4 video of 29 seconds length and 31.2Mb size. Here are the results.

Platform
VMware
Virtualbox
Macintosh
Convert 1280×720 Mpeg4 to 1280×720 Divx3
17
16
15*

* – Mac filter used one core

These were more what I was expecting. VLC only uses one core of the Mac’s Core 2 Quad processor so the results are fairly close. Using either virtualisation package gives performance close to the Mac, so for single tasks they are a good solution for those Windows applications you have to run. Of course the Mac would perform far better when multiple applications are running, but for that you could use boot camp to run Windows XP natively and benefit from the full four cores available.

So which of these packages would I choose? Virtualbox is free so has a big advantage over the commercial VMware. The downside is that Virtualbox is still beta software, and doesn’t support direct file dragging from host to guest OS. The gap between Beta 2 and Beta 3 on the Mac was eight months, so hopefully the time to the final release won’t be as long.

Advertisements
  1. Abu Yahya
    December 2, 2008 at 2:46 pm

    Thanks for the wonderful analysis. There’s the full version of VirtualBox available now, so I think it would be a great idea to try the comparison again.

  2. basshead
    December 2, 2008 at 3:14 pm

    I’ve already posted part one of an updated comparison of Parallels, VMware Fusion and Virtualbox as Mac VM’s – November 2008 Part 1. Further parts will be posted when completed.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: