I hadn't cleared the cache first, but I just did by running:

checkpoint;

go

dbcc dropcleanbuffers;

go

Then ran the query again. Completed in 5:25.

I did not update statistics, I will do this and repost the results.

The hardware shouldn't be too taxed, each host has 128 GB ram, and the host this VM is currently on only has 3 other windows 2003 servers running on it. I wasn't really thinking the vCPU would help, but I expected the RAM to help a lot.

How can I go about checking the IO throughput?

Thanks