Vista Native vs Windows 7 in Virtualbox

Just out of curiosity I ran Windows assessment of Windows 7 running in virtualbox (with VT-x enabled) and Vista. Here is a screen shot of the comparison.

Vista Native vs 7 on Virtualbox 

What was amazing is the Hard disk Score, Virtualbox HDD is faster than the native disk !! Processor score is not too shabby, but I expected more with VT-x turned on. So how much does VT-x have influence on the Virtualbox performance? I tried to do the same score with VT-x turned “OFF”.

To my surprise, I got the same score! Perhaps, the windows assessment tool does not run any ring-0 code ( as far as I know, VT-x helps better management of ring-0 code).

Anyhow I’m posting the screen shot of the windows 7 on virtualbox with VT-x turned off.

Seven_virtualbox_without vt-x

If Anybody has a better idea as to why VT-x failed to change the scores in windows 7 assessment, please feel free to comment 🙂


6 thoughts on “Vista Native vs Windows 7 in Virtualbox

  1. I have an idea… VT-x is supposed to virtualize instructions on the host processor so virtualbox doesn’t have to emulate them. Even if things move slower in real time, maybe the windows benchmark didn’t see it because it sees the world in emulated time. If it takes the same number of fake cpu cycles, whether emulated or virtual, the guest OS would see the same result.

    Or maybe the windows experience is going by the precessor’s specs… clock speed and cache, etc.

  2. Thats a good theory. So Windows could be looking at the processor time, which could be a bit skewed with virtualization. However virtualized environments simply cannot afford to skew wall clock time. This is a very important point to consider. Because wall clock time is very important for many applications, they would have ensured its consistancy.
    However, it could also be the specific nature of tests, the instructions that Windows 7 uses.
    However, I dont think that the They go by processor specs and clock speed. That would imply that windows 7 should keep track of every processor and its variant in its database (Possible, though unlikely).

  3. Thanks. Besides the wall clock, which actually does become skewed over long periods of time (but is not intentionally skewed), there are 3 clocks which can be used.

    One is the cpu tick count, which naturally moves faster or slowed on frequency scaled cpu’s. This one is the easiest to use in measurements, but the least reliable. Another is a dedicated hardware performance clock, which is often used in games and performance-critical apps. Both of these are high-resolution. A third clock is closely related to the wall clock, and it ticks about 18 times a second IIRC.

    If my previous idea is correct, then it’s probably using the cpu clock or the hardware clock. That could also skew the hard disk performance results, since the drive will be moving faster relative to the slower clock.

    Thanks for this very interesting blog.

  4. I don’t know if you know this, but the only thing you need to take into account here is PROCESSOR, you get 5.2 in native and 4.3 in virtualization, it’s a little lower, but looks normal to me, you are running to Win7 OS, one inside another, one thing is odd is that if you disable virtualization, the scores remains the same.

    Something wrong is going on here.

    You get 1.0 score, because the graphic subsystem is emulated & i think if you don’t enable AERO you will always get 1.0 + you can’t enable AERO because VIRTUALBOX has not his drivers certified.

    Wait for better drivers for the guest os, and you will se you will get a lot higher score.

Leave a Reply

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

You are commenting using your 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