John Pritchett and I had a discussion going on how to improve TWGS performance even more. My theory was the hard drives may be playing a big role in overall system speed, he disagreed. So I setup a RAM drive (meaning an emulated hard drive that is based ENTIRELY in memory), to see if it helped. The purpose was to eliminate the slow access time of the hard drive from the equation. To my surprise, a system with a RAM drive was SLOWER! This is completely opposite to any other benchmark ever seen with a RAM drive v.s. Hard drive. All this means is that the TWGS simply doesn’t access the hard drive often enough for it to play a significant role in performance. Windows already does a great job of caching things in memory, making a RAM drive irrelevant. Look at the results..

NO Ramdrive
————–
1 players running colos – 1.36 minutes
2 players running colos – 2.22 minutes
3 players running colos – 4.40 minutes

With Ramdrive
—————
1 players running colos – 1.32 minutes
2 players running colos – 2.27 minutes
3 players running colos – 4.44 minutes