VirtualBox Benchmarks !!!
-
- Volunteer
- Posts: 3342
- Joined: 10. May 2007, 16:59
- Location: Israel
VirtualBox Benchmarks !!!
At last !
I have found time this weekend to make some serious benchmarks with many contenders...
VirtualBox 1.5.0, KVM, VirtualPC, VMware, Qemu, and of course... Real Hardware !
UPDATE: Good news !
I have found a permanent location for my performance benchmarks results:
http://www.violtan.com/ae/KVM-VirtualBo ... hmarks.xls
For questions email me at al4321 at gmail.com
-Technologov.
I have found time this weekend to make some serious benchmarks with many contenders...
VirtualBox 1.5.0, KVM, VirtualPC, VMware, Qemu, and of course... Real Hardware !
UPDATE: Good news !
I have found a permanent location for my performance benchmarks results:
http://www.violtan.com/ae/KVM-VirtualBo ... hmarks.xls
For questions email me at al4321 at gmail.com
-Technologov.
Last edited by Technologov on 26. Jan 2008, 02:30, edited 1 time in total.
-
- Posts: 44
- Joined: 22. Jun 2007, 23:21
- Contact:
-
- Volunteer
- Posts: 3342
- Joined: 10. May 2007, 16:59
- Location: Israel
OK, I have uploaded it again: (the link will expire soon)
Link: http://download.yousendit.com/910490EE0564BB5C
newhere123: I have updated the results to include Parallels, according to your request.
Link: http://download.yousendit.com/910490EE0564BB5C
newhere123: I have updated the results to include Parallels, according to your request.
-
- Volunteer
- Posts: 3342
- Joined: 10. May 2007, 16:59
- Location: Israel
OK, I have uploaded it again: (the link will expire soon)
http://download.yousendit.com/EBC067BC56A2DC40
I became too lazy lately... but I really need to update some of those benchmarks... (as well as make a permanent link)
http://download.yousendit.com/EBC067BC56A2DC40
I became too lazy lately... but I really need to update some of those benchmarks... (as well as make a permanent link)
-
- Volunteer
- Posts: 3342
- Joined: 10. May 2007, 16:59
- Location: Israel
Good news !
I have found a permanent location for my performance benchmarks results:
http://www.violtan.com/ae/KVM-VirtualBo ... hmarks.xls
-Technologov
I have found a permanent location for my performance benchmarks results:
http://www.violtan.com/ae/KVM-VirtualBo ... hmarks.xls
-Technologov
-
- Volunteer
- Posts: 3342
- Joined: 10. May 2007, 16:59
- Location: Israel
Future benchmarks
================
1. someone asked me to benchmark Emacs - Unfortunately, I don't know Emacs, so I cannot benchmark that one. If you have benchmark scripts for Emacs, I would be glad discussing it with you.
2. I plan to re-test Passmark, as I have a new build that runs more stably across all platforms. (esp. Qemu)
3. Windows XP - I will setup it automatically, via unattended setup, instead of manual.
This will negate human-error factor, and will eliminate my slow CD-key typing speed.
4. I wanted to test LGuest, but openSUSE 10.3 doesn't support it, and I'm too lazy to build custom kernels.
5. I will write near each test, if I used internal or external timer. (internal=in guest) It may be inaccurate.
6. any other ideas?
-Technologov
================
1. someone asked me to benchmark Emacs - Unfortunately, I don't know Emacs, so I cannot benchmark that one. If you have benchmark scripts for Emacs, I would be glad discussing it with you.
2. I plan to re-test Passmark, as I have a new build that runs more stably across all platforms. (esp. Qemu)
3. Windows XP - I will setup it automatically, via unattended setup, instead of manual.
This will negate human-error factor, and will eliminate my slow CD-key typing speed.
4. I wanted to test LGuest, but openSUSE 10.3 doesn't support it, and I'm too lazy to build custom kernels.
5. I will write near each test, if I used internal or external timer. (internal=in guest) It may be inaccurate.
6. any other ideas?
-Technologov
First of all, I'd like to thank you for your work.
It's really interesting to compare all these systems.
As requested, I have some more ideas for you.
I think it would be also interesting to compare all your figures against these other systems, which are totally free:
- Microsoft Virtual Server 2005 R2;
- VMWare Server 1.0.4;
- VMWare Server 2.0 beta.
Testing some other applications like Emacs is also a good idea. You could measure the time to launch Emacs with fni's ".emacs" file and see how it compares on all these systems...
I know it's a lot of work but could you try that?
Thanks a lot,
-fpz
It's really interesting to compare all these systems.
As requested, I have some more ideas for you.
I think it would be also interesting to compare all your figures against these other systems, which are totally free:
- Microsoft Virtual Server 2005 R2;
- VMWare Server 1.0.4;
- VMWare Server 2.0 beta.
Testing some other applications like Emacs is also a good idea. You could measure the time to launch Emacs with fni's ".emacs" file and see how it compares on all these systems...
I know it's a lot of work but could you try that?
Thanks a lot,
-fpz
-
- Volunteer
- Posts: 3342
- Joined: 10. May 2007, 16:59
- Location: Israel
For the Windows guests, you can install and use EmacsW32 (see http://www.ourcomments.org/Emacs/EmacsW32.html, patched Emacs).
Once installed, upon start up ("emacsclientw.exe"), it will load an init file and follow the instructions there (telling him to load more packages).
Set an HOME environment variable to some directory (for example, "C:/emacs") and put in there any "dot emacs" file that you would use as representative for your test.
Mine is available on http://www.mygooglest.com/fni/dot-emacs.html#sec4.
Even without modifying anything (I reference some directories where I put my extra packages), it should be sufficient to load and customize the default stuff (available with the default installation you've made above).
At the end of the launch, it outputs (in the so-called minibuffer area) the amount in seconds it took to launch itself. That's the information you're interested in.
fni
PS- If you have difficulties creating a ".emacs" file (not very Windows-like naming), you can name it "_emacs". Emacs searches for both names when starting up...
Once installed, upon start up ("emacsclientw.exe"), it will load an init file and follow the instructions there (telling him to load more packages).
Set an HOME environment variable to some directory (for example, "C:/emacs") and put in there any "dot emacs" file that you would use as representative for your test.
Mine is available on http://www.mygooglest.com/fni/dot-emacs.html#sec4.
Even without modifying anything (I reference some directories where I put my extra packages), it should be sufficient to load and customize the default stuff (available with the default installation you've made above).
At the end of the launch, it outputs (in the so-called minibuffer area) the amount in seconds it took to launch itself. That's the information you're interested in.
fni
PS- If you have difficulties creating a ".emacs" file (not very Windows-like naming), you can name it "_emacs". Emacs searches for both names when starting up...