I have spoken to a concerned member of the discord and heard that there were potential FPS issues caused by SAIN 5.2 (latest testing).
I decided to test it using a limit on how much AI can spawn (AI amount: medium) and the Tarky tool from SSH to simulate a worst-case scenario with a shitload of AI doing a lot of things at once, right in front of the player (teleport-to-crosshair)
I have conducted this test on the following specifications:
RX 6700XT
16GB of DDR4 3200MHZ
Ryzen 7 5800x3D
The FPS was limited to 100 for the sake of heat and freesync. This was tested on customs daytime.
The AI view distance has been set to 100m for the sake of consistency with the radar (also set at 100m)
Additional controlled variables: AI limit uninstalled to ensure consistent results; configs reset for the sake of fair testing on clean copies.
Here are the results:
No-SAIN:
-100FPS with no AI within 100m
-90-95FPS with a few active AI (10-12AI counted over 3 sessions)
-78-85FPS when all 20 decide to fling shit at each other in the immediate vicinity (10m)
SAIN 5.2:
-100FPS with no AI within 75m
-75-80FPS with less than 10AI active (the worst one was with 2 AI that I accidentally encountered at the gas station post-testing, and it dropped like crazy)
-50-80FPS (very variable frame times) when all AI are engaged in shit flinging.
Since this was only a personal test, it shows nothing 100% conclusive, but given the current results and the controlled setting, it would be reasonable to conclude that there may be a cause for investigating the performance of the mod in the future.
The mod list used:
-Amands graphics
-Game hud
-DoorBreacher
-Fershite balance patch
-FragmentationDisabler
-Waypoints
-Bigbrain
-Radar
-FU inertia
No errors in log.
Sorry for the long post, here's a potato to comfort you. gives potato
I have spoken to a concerned member of the discord and heard that there were potential FPS issues caused by SAIN 5.2 (latest testing).
I decided to test it using a limit on how much AI can spawn (AI amount: medium) and the Tarky tool from SSH to simulate a worst-case scenario with a shitload of AI doing a lot of things at once, right in front of the player (teleport-to-crosshair)
I have conducted this test on the following specifications:
RX 6700XT
16GB of DDR4 3200MHZ
Ryzen 7 5800x3D
The FPS was limited to 100 for the sake of heat and freesync. This was tested on customs daytime.
The AI view distance has been set to 100m for the sake of consistency with the radar (also set at 100m)
Additional controlled variables: AI limit uninstalled to ensure consistent results; configs reset for the sake of fair testing on clean copies.
Here are the results:
No-SAIN:
-100FPS with no AI within 100m
-90-95FPS with a few active AI (10-12AI counted over 3 sessions)
-78-85FPS when all 20 decide to fling shit at each other in the immediate vicinity (10m)
SAIN 5.2:
-100FPS with no AI within 75m
-75-80FPS with less than 10AI active (the worst one was with 2 AI that I accidentally encountered at the gas station post-testing, and it dropped like crazy)
-50-80FPS (very variable frame times) when all AI are engaged in shit flinging.
Since this was only a personal test, it shows nothing 100% conclusive, but given the current results and the controlled setting, it would be reasonable to conclude that there may be a cause for investigating the performance of the mod in the future.
The mod list used:
-Amands graphics
-Game hud
-DoorBreacher
-Fershite balance patch
-FragmentationDisabler
-Waypoints
-Bigbrain
-Radar
-FU inertia
No errors in log.
Sorry for the long post, here's a potato to comfort you. gives potato