I'm currently on putting the finishing touches on my new computer. I used the excellent guides provided by Jeff Atwood (Parts I, II, III, and IV) and some, albeit severely diminished, historical knowledge to build the PC from components. A few weeks ago I posed the question of dual-core or quad-core. Jeff left a comment on the post:
"Do you *really* expect to be using this same machine 2 to 4 years from now?
I think that's a stretch.. buy for the next year, and let the future sort itself out later. You always end up upgrading the mobo / memory / CPU together anyway."
Though his comments carry a lot of weight, I decided to not heed Jeff's advice and to run with the quad-core processor. As much as I hope to buy new hardware more often, my own history would prove otherwise. I simply don't buy new hardware that often, despite Jeff's warning.
There is one other reason I opted for the quad-core. I want to set this PC up to use virtual machines for development. I was bitten last week by having Visual Studio 2005 and Visual Studio 2008 Beta 2 installed on the same machine. In a virtual environment I could have a 2005 virtual machine and a 2008 virtual machine. Having four cores then allows me to assign a virtual PC to run on specified cores.
Will the choice to go quad-core pay off? We'll never know, cause I don't purchase hardware often enough to buy a dual-core to compare. I'm curious about the drop in performance between a host machine and a virtual machine. I ran some benchmarks on the new machine running Windows XP on the host and will run the same benchmarks with Windows XP running as a virtual machine. I will report back when I have more data.