Crysis Benchmark: Low FPS. WHY?

If you tried my Crysis benchmark you got low FPS at Very High settings. Yes, it's true. I spent much time with find the more stressed graphic settings for the benchmark. So if you set the Very High option let you remember that it's a tweaked setting so do not be surprised that you will be get 15-25 FPS on 1280x1024 (on NVIDIA 9800).

Maybe in the future I published these settings.
Oh, I forgot: DO NOT BELIEVE the big words like: "Ultra High Quality settings cfg, download it!").

Get errors because the game type is different...

I get many emails about errors as:

"Hi, I bought my game via Steam, but your benchmark isn't support it. Help me!"

Ladies and Gentlemen: I do not have money that let me buy all games on DVD and via Steam, EA, etc... This is my hobby and I honour the games developers with my money once (buy an original copy), but that's all. If it is important to you buy me a copy of this.

Updated Far Cry Benchmark

My very old, but favourite benchmark step in the new era. I fixed some minor error, the results are readable and added some of the new techniques. Release: 03. 06 . 2008

Some datas about HardwareOC Far Cry benchmark...

• More than 820.000 download (on 9 sites)
• The first version published on 20. 07. 2004
• During this time I got more than 6000 emails about this benchmark

I think this is the last version, so thank you this four year.

Next in the benchmarks...

• I would like to add it to all the chart function (DONE).
• Multilingual support.
• Steam compatibility for Enemy Territory: Quake Wars benchmark.
• Steam compatibility for Doom 3 benchmark.
• Update Quake 4 benchmark (compatibility for v1.4.2 patch).
• CPU-Z and GPU-Z validation.
  
• And try to do it against optimizations. My bench tries to rename the executable files to randomly (DONE).

Optimizations

Optimizations... Adjusting the result to the natural world. To your vga card. I hate that I am fooled always. But we know it, everything is about the money.

So this world is here and it may be yours. All the same, what there is in your machine. ATI, MATROX, NVIDIA, S3... all the same. Just do not believe in the scores. We usually purchase faster and faster CPUs and graphics controllers not to get higher scores in synthetic benchmarks (like 3DMark, Aquamark etc.), but to have the best visual and general quality in real games that we are about to play for days or weeks. Certainly this does not mean that we question the reason for existence of synthetic products, but they should be used for examining a hardware’s theoretical capabilities. Getting higher scores in them must not significantly influence which product we are about to buy. Synthetic benchmarks are static. They always render the same scenes from the same angles with the same settings. So it is possible to create a device driver which gets advantage of this stativity, therefore it makes the hardware do better under these circumstances than how it could perform under real applications. Never choose a graphics controller just because it gets more scores in a well-known synthetic application.

So... welcome to the world of real benchmarks!