VMware VS Parallels SCREENSHOT CPU usage

Look at the difference and yes they both have the same app installed. Maybe parallels has a little less stuff on it but they are dang close and these are fresh installs

Picture1.png
 
Last edited:
Number one you are not looking at the correct process my friend, the actual executable of VMware which uses proccessor cycles is called vmware-vmx and it's owned by root, which means that you can see it in Activity Monitor under "All proccesses".
The one you are seeing is just the VMWare console app which uses no cycles.

Number two, if you want Parallels to be around 9% when idle, and 2% instead of 10% when paused, go in the preferences of your VM file and set the "input sound device" to "null".
 
Last edited:
I can say from everyday usage that vmware-vmx is far more efficient idle then parallels. Parallels does not appear to get the "idle" concept. It smells like they are doing a lot of polling.
 
pwith said:
Parallels seesm to be using a helluva a lot memory in comparison to vmware - 220mb to about 20mb

Yes it does appear that way. Again, as was mentioned yesterday, the screenshot does not include the actual vmware guest or the overhead used with it. vmware-vmx is a root program and cannot be seen the way websyndicate set up activity monitor.
 
Yeah you need to change your drop down filter from "My Processes" to "All Processes" which requires a root authentication.

I realize you're trying to make VMWare look good, but your lack of OSX knowledge makes you lose your credibility.

I've recently switched from Parallels to VMWare and I'm happy about the enhancements they're making. VMWare has been in the business far to long to not make a good product. I really like that I can use my VMWare VMs in any of their products and that their VM Tools work in Windows and Linux. At this point its personal preference, until one product out-shines the other.
 
crazibri said:
Yeah you need to change your drop down filter from "My Processes" to "All Processes" which requires a root authentication.

I realize you're trying to make VMWare look good, but your lack of OSX knowledge makes you lose your credibility.

I agree with that 100%
 
Stefdar said:
Number one you are not looking at the correct process my friend, the actual executable of VMware which uses proccessor cycles is called vmware-vmx and it's owned by root, which means that you can see it in Activity Monitor under "All proccesses".
The one you are seeing is just the VMWare console app which uses no cycles.

Number two, if you want Parallels to be around 9% when idle, and 2% instead of 10% when paused, go in the preferences of your VM file and set the "input sound device" to "null".
Nice tip re: the input sound device. Thx! It only dropped me a few percentage points, but every bit counts.
 
Here are my comparisons

Here are Parallels and VMWare Fusion (the latest of each,) running XP Professional SP2. As you can see, both are at 'idle' inside Windows, and have been so for some time. This assures a more even comparison. Both are configured with 512 MB of RAM.

The first comparison shows Parallels configured for 'normal use', and VMWare configured to be the same as Parallels (only one processor, 3D acceleration turned off,) while the second shows VMWare configured for 'normal use' (two processors, 3D acceleration on.)

This is running on a MacBook Pro 2.0 GHz Core Duo with 2 GB of RAM.

Screen Shot 1, Screen Shot 2.

As you can see, VMWare uses significantly less processor time when idle when it is configured the same as Parallels. When it is configured with all the extras on, though, it uses the same.
 
Also worth pointing out in those screen shots that VMWare seems to be using quite a bit more real memory, which seems odd. I guess Parallels delays allocation of the ram until requested?
 
websyndicate said:
Look at the difference and yes they both have the same app installed. Maybe parallels has a little less stuff on it but they are dang close and these are fresh installs
This is not a fare comparison.
First of all, you're looking at the wrong process - the actual work is happening in the different process running under root privileges - wich is potentially more dangerous for your system in the case of its crash.
You need to add the memory and cpu usage of all processes used by the virtualization system together to get the right number.

Please, analyze this problem with some more detailes and you will see that situation is not that bad if you compare Parallels to the competitors :)

Anyway, competition is definitely a good thing ;)
 
VM vs Parallels

I tried VMWare and have to say it's not ready for primetime. I didn't compare CPU usage, but there was no comparison: under VM, my HDD was going berserk. I think it doesn't handle memory as well as Parallels does. Bottom line, I simply couldn't use it. I keep a large number of programs open on WinXP (Most of OFfice2007, plus stock trading software) and switch back and forth between that and OSX. VMWare was just slow and constantly writing to disk.
 
Oh, and in my comparison, one thing I noticed is that, as websyndicates's screenshot also shows, VMWare uses up more memory; and Windows in the VM takes noticdably longer on VMWare than Parallels.
 
I don't have any metrics, but VMWare seems to have much fewer blocking operations. VMWare may be using more CPU and more Memory. It also takes longer to boot my bootcamp VM than Parallels does. However, while the Bootcamp vm is booting in parallels, my computer is worthless. It is unable to open Applications, clicking already open applications stalls for some time with the spinning beachball of doom. With VMWare, it takes a very long time to launch the VM, but at least the computer responds to other I/O requests while I have to wait.
 
From what it sounds like VMWare boots faster and Parallels runs faster... VMWare is BETA and things are finally stable at our office with Parallels so I don't think I'm going to switch.

Maybe Parallels can take a page out of VMWare's book and get faster booting.
 
Back
Top