Basically, your results demonstrate the fact that Viper Racing is CPU-limited on your particular hardware configuration. In a sim like Viper Racing, there is a tug-of-war between the amount of resources devoted to rendering the action versus the resources devoted to generating the action. Viper Racing and other sims that have a number of AI-generated opponents have to allocate substantial CPU cycles to those AI opponents. If you are viewing a race replay, Viper Racing is using zero CPU cycles for opponent AI whereas during a race, it is expending a considerable portion of available CPU cycles on opponent AI. This equates to a higher frame rate during race replays, assuming you are not hitting a frame rate upper bound on your graphics card or a frame rate limit imposed by the developers.

If the Viper Racing benchmark is not participatory (you as a user don't play during the benchmark), then it sounds like the paths of the vehicles in the benchmark are predetermined, so zero CPU cycles are spent on AI and you see frame rates similar to an after-race replay. It is also possible that the number of vehicles in the benchmark is less than then typical number in your races, so fewer CPU-cycles would be required for AI-generation and higher frame rates would occur.

akula65 wrote on 2021-10-21, 09:55:

Basically, your results demonstrate the fact that Viper Racing is CPU-limited on your particular hardware configuration. In a sim like Viper Racing, there is a tug-of-war between the amount of resources devoted to rendering the action versus the resources devoted to generating the action. Viper Racing and other sims that have a number of AI-generated opponents have to allocate substantial CPU cycles to those AI opponents. If you are viewing a race replay, Viper Racing is using zero CPU cycles for opponent AI whereas during a race, it is expending a considerable portion of available CPU cycles on opponent AI. This equates to a higher frame rate during race replays, assuming you are not hitting a frame rate upper bound on your graphics card or a frame rate limit imposed by the developers.

If the Viper Racing benchmark is not participatory (you as a user don't play during the benchmark), then it sounds like the paths of the vehicles in the benchmark are predetermined, so zero CPU cycles are spent on AI and you see frame rates similar to an after-race replay. It is also possible that the number of vehicles in the benchmark is less than then typical number in your races, so fewer CPU-cycles would be required for AI-generation and higher frame rates would occur.

Oh, interesting. Thanks for the enlightenment.

Patching the game might give some performance improvements:
https://www.patches-scrolls.de/patch/4674/7/

There's a later unofficial patch available for Viper Racing (1.2.4) beside the official patch 1.1 - both are available on my website (-> https://soggi.org/misc/game-patches.htm).

kind regards
soggi

soggi wrote on 2021-10-22, 09:00:

There's a later unofficial patch available for Viper Racing (1.2.4) beside the official patch 1.1 - both are available on my website (-> https://soggi.org/misc/game-patches.htm).

kind regards
soggi

Cool, thanks. According to PCGamingWiki it also removes limits, though, which idk would be a good thing on this system. I've used your site quite often, that CART Precision Racing glide patch ws super useful. Runs so much better than the stock Direct3D driver.

Nice to hear, you're frequently using my page! If I had more time for my project, there would be more exotic patches like the CART Precision Racing one. I've got a big heap of things (drivers, BIOSes, game patches) in the pipe, but actually not enough time. Hopefully it's getting better soon...

Did the Viper Racing patches help you!?

kind regards
soggi