http://http.download.nvidia.com/down...VBenchmark.zip
Slipped in under the radar, was mentioned in passing in the hardware thread but I'm almost positive its brand fucking new. Offers both 1080p and 720p settings, though custom settings (AA and what not) are not available. Plays several rendered scenes found in the Alpha test. I will be making a leaderboard in this original post and keeping track of submitted scores, since people are always trying to compare systems and this will save endless page scrolling.
For a legitimate score to be added to the leaderboard, you must screenshot and attach the image with both your score and system specs viewable and mention if it is low or high.
Also, if you're using an ATI card, try to mention your cards specific model as it only says the series in the bench! (5850 vs 5870, if you don't tell me I'm going to guess)
SLI/CROSSFIRE DOES NOT APPEAR TO BE FUNCTIONAL, YOU WILL ONLY BE USING A SINGLE GPU CORE IF YOU RUN WITH IT ENABLED RESULTING IN LOWER THAN EXPECTED SCORES ALL SLI/CROSSFIRE SETUPS AND CARDS LIKE THE GTX295, HD5970, HD4870x2, ETC
High Settings Leaderboard
http://img191.imageshack.us/img191/3927/highl.jpg
Low Settings Leaderboard (Disclaimer: Running at these settings is going to be largely effected by your CPU, making graphics card mostly irrelevant)
http://img704.imageshack.us/img704/4440/lowy.jpg
Last update: 6/17/10, currently accurate to post #425
Analysis: Even though it says The Way It's Meant to be Played stamped across the top, GPU manufacturer thankfully does not seem to be a factor. When comparing similiar GPU/CPU configurations (Saulin/Killerc, Tri/Kuro, etc), the scores are pretty identical and in-line with what you would expect from the varying clocks. CPU seems to be a larger than average(compared to other games/benches) factor at high settings, and completely dominates low settings to the point where GPU hardly matters at all.
How the benchmark relates to current beta performance
The following tests done on a system that scores 4500 High/5600 Low. Benchmarked with FRAPS logging the Min/Max/Avg framerates, in an instanced location to reduce performance discrepancy. Travel loops will be reproduced identically to the best of my ability and will not be overly complicated. More trials may be added in different locations at a later date.
Trial presets were determined by me, there is no 'minimum'/'maximum' option you can choose. Buffer Size was Resolution for all tests, Depth of Field being ignored.
'Minimum' -1280x800, Shadows Lowest, Multisampling Off, Ambient Occlusion Disabled, Texture Quality Low, Texture Filtering Lowest
http://www.razorbites.com/celeras/minimum.jpg
'Standard' -1920x1200, Shadows Standard, Multisampling 4x, Ambient Occlusion Disabled, Texture Quality Standard, Texture Filtering High
http://www.razorbites.com/celeras/standard.jpg
'Enthusiast' -1920x1200, Shadows Highest, Multisampling 16x, Ambient Occlusion Disabled, Texture Quality High, Texture Filtering Highest
http://www.razorbites.com/celeras/enthusiast.jpg
'Maximum' -1920x1200, Shadows Highest, Multisampling 16xQ, Ambient Occlusion Enabled, Texture Quality High, Texture Filtering Highest
http://www.razorbites.com/celeras/maxed.jpg
NDA is down, screenshots are allowed. Basically just walked a loop in the starter room on the varying settings with fraps benchmarking FPS.
FINAL FANTASY XIV ©2010 SQUARE ENIX CO., LTD. FINAL FANTASY is a registered trademark of Square Enix Holdings Co., Ltd. All material used under license.