VT performance benefits - what's the deal?

F

fizzer fits

Bit poster
Hey all,

I'm getting a new laptop (Windows - can't afford Mac... ) with an Intel Core Duo processor.

I have two choices in running VMs;

VMWare, which isn't too customizable and doesn't really support VT but is free (as in beer)...

And Parallels, which is pretty robust and supports VT, but costs 50 bucks ( a great price, but all of my money went to that laptop).

So how much of a performance bump does VT provide? Has someone run Parallels 2.1 w/ and w/o VT enabled in the bios and noted the performance difference? 5%? 10%? 100%?

Any official word from the parallels team as to how much VT improves performance on the guest and host OSes?

Much appreciated
 
VT provides a pretty impressive performance boost. In real-life testing, we've noticed as much as a 150% performance gain in VT powered machine versus a non-VT powered machine.

For example, on my MacBook Pro, my Windows XP virtual machine actually runs FASTER than the ThinkPad I used for the past year! That performance is largely due to our hypervisor-powered virtualizastion engine, and the boost that VT (included on all new Macs) provides.

My advice would be to build a VM using the free 15-day trial of Parallels Workstation. Once you see how fast things run, I think you'll want to own a copy!
 
That, sir, was exactly the type of responce I was hoping for - I'll most assuredly pick up the software after the laptop comes in.

There is nothing on the Internet right now that discusses VT in any practical manner, especially benchmarking. If I have time I'll make sure to run some tests if the laptop's bios permits me turning VT off easily; I'll be happy with anything more than 10% as it'll relieve some of the buyer's regret from getting the laptop.

Anywho - VMware people aren't *really* implementing it yet except for their 64bit server applications I think (their forums, if I read correctly, have the devs saying that if anything VMPlayer is hobbled with VT on), and if other VM programs use it at the moment they definately aren't saying so. If someone pubishes some extensive benchmark comparisons with VT on and off they'll be the only resource that I have spotted on the web which does so. If I have time, maybe I will - but I definately don't have much time right now.

Thanks a million for the response!
 
You must log in or register to reply here.
Back
Top