I did some benchmarks awhile back. Just Case 2, Tomb Raider, Metro Last Light and War Thunder. I wish I saved the numbers so I could give you something more concete, but from what I remember, all of them ran between 97-99% of what they did on baremetal.
Edit: Actually I did. Now I only ran each of these once. These are just the average FPS JC2 89% War Thunder 96.5% Tomb Raider 95.5% Metro LL 105%
I'd like to go back and check the Metro one, cause something doesn't seem right there.
I was hoping you read the part where my CPU doesn't support virtualization:/ we were discussing other ways to implement Linux into my computer while still being able to game and use Internet explorer. Thanks for your input though!
Yes, that should work. Once installed and the windows drive reattached you will be able to edit grub to add windows as a boot option, so no need to enter the BIOS.
Having said that I've never had any problems with grub if I set up the new disk as disk to boot from, and make sure grub writes itself to that one.
The performance improvement is possible, it could also be that your settings were not 1:1 between the bare metal and KVM installs.
How a performance improvement is possible is down to how things are getting scheduled on the procesor. In effect KVM/Linux is optimising how the threads from the VM are getting executed on the physical host system. I doubt you would ever get anything you can notice (others may disagree) but it can certainly be measured. You can get a similar benefit on bare metal windows by adjusting processor affinity and task priority - especially when you prioritise your game process ahead of all the bloat that is often found on windows machines.