Upgrading Geordi
to
It had long been my plan to give Geordi a mid-life upgrade by GPU. One step that will do a lot, and is a lot simpler to plug-out and plug-in than mucking with something like a replacement CPU--that can wait some more years for a totally new machine. I'd even for a long time had a general idea of what I was going for. To make the transition as possible, go from the same family member from one generation to another. So since I had a GeForce GTS 450, I figured it would probably be something like a 650. Doing a bit of research recently, though, I see the (number)50 is relatively a bit on the lower end this time. The performance jump wouldn't be greater, but it would use a lot less power and whatnot. While it was an objective to pick something that wouldn't be hot and noisy that seemed like giving too much up, so I went with the next half-step up: the GTX 650 Ti. Compared to the GTS 450 it has about twice the processing capability and uses barely more power, and the specific model I went with has twice the RAM (2 GB, up from 1). Decent upgrade.
I thought it would be interesting to go back to the benchmarks I ran when upgrading the PC's stock card to GTS 450 over two years ago and run them now on this even newer equipment. I also tossed in one new comparison using one of the benchmarks built into Just Cause 2. One thing that makes the comparison somewhat less even is back then I was using the iZ3D driver for 3D mode, while now I'm using NVIDIA's 3D Vision.
Mitchell Geordi
w/ Radeon 5450 Geordi
w/ GeForce GTS 450
2D Geordi
w/ GeForce GTS 450
3D Geordi
w/ GeForce GTX 650 Ti
2D Geordi
w/ GeForce GTX 650 Ti
3D
FF XI "High"
1024x768 1877 5942 9293 5234 7226 N/A
FF XIV "Low"
1280x720 ~0
(Sound only) 716 3199 1521 3799 2934
FF XIV "High"
1920x1080 Didn't
bother 351 1708 736 3290 1830
Just Cause 2
"The Dark Tower"
1920x1080 N/A N/A 35.27 fps 17.93 59.22 32.29
Final Fantasy XI is a bit of a weird one. I expected that something made for the machines of a decade ago, doesn't use modern effects, and maxes out at 1024x768 wouldn't see much gain... but it actually seems to be lower? I did run it a couple times to be sure. Also, no 3D score for this one because I guess this benchmark doesn't work with 3D Vision.
Final Fantasy XIV and Just Cause 2 were mostly what I expected--results about doubled. A few exceptions are the lowest shown cases for each. I think at the less taxing times they were hitting frame rate caps, so the average improvement appeared to be less. The highest resolution 3D one for FF XIV also saw an extra large boost--though it's weird in that it only worked out that way the second time I ran it. The first time I ran it I thought the result of 1198 seemed unusually low, so I gave it another go. I guess something in the background was distracting the computer the first time around?
Beyond these benchmarks the only real-world test I've put it through yet is Skyrim--but since it detected the new card it modified the settings to suggested defaults, so I'll have to fiddle with that again. BUT thanks to the extra RAM I'm now able to use the official high res texture pack without it chugging along like an overencumbered character.
On a practical note, this card should stay a little cooler because I made a doofus mistake in 2010. Replacing the small included card with the bulkier GTS 450, it took a lot more height-- but I didn't notice that it was designed to use two of the rows on the back of the case unlike the included card, so the vent was mostly blocked by a thin piece of metal. Though the GTX 650 Ti is smaller overall, it has outputs on both the top and bottom parts that stick out slightly. It thus couldn't fit in without me figuring out which part needed to be removed and should have been 26 months ago.