The command used for this was:
hdparm -Tt /device
Not even sure if this is a really great way to benchmark drives, but here are the results:Digital Ocean where my drive was /dev/vda:
Timing cached reads: 17246 MB in 2.00 seconds = 8632.56 MB/sec
Timing buffered disk reads: 892 MB in 3.00 seconds = 297.20 MB/sec
Next was Linode where my drive was /dev/xvda: Timing cached reads: 12680 MB in 1.99 seconds = 6381.79 MB/sec
Timing buffered disk reads: 614 MB in 3.00 seconds = 204.66 MB/sec
Was expecting a more dramatic difference considering DO uses SSD's and Linode does not (I am not in the SSD beta). When you compare consumer grade HD's and SSD's the difference is night and day.I do not have results from my Poweredge because it's currently powered off. The fucking thing sounds like a C5 doing an assault landing on a tattered runway during a hail storm.
Take this all with a grain of salt just sharing since we're comparing hosts here.
Finally, I love Linode and have never had any shit with them. Digital Ocean is also great although I have only really dipped my toes into their stuff. Have not tried their recently launched internal-networking yet. Also, unfortunately you cannot (easily) downsize a box, only expand one. Rather you need to create a backup and then a new box and restore the backup to that new box or something. Seems silly. Linode lets me do this, and I do it for some ghetto on-demand scaling from time to time.
Sidenote: If you wanna muck around with FreeBSD I had some good experience with doing that on Rackspace. Doing it on EC2 is not as rewarding.