Contains executable files or external scripts. Ensure you trust the creator and apply your own security measures.
An updated version of the CPU Performance Patch for Virt-A-Mate 1.22.0.13
Included - everything is set up to work together:
Main changes:
Requirements and installation:
The included settings were tested on my Ryzen 7 5700X3D. They are not a universal preset.
For a different CPU, please check your core and logical processor counts, then adjust and test the settings on your own system. More threads do not automatically mean better performance, and there is no single formula that gives the best settings for every CPU and scene.
I’m also working on a much bigger 3.0 update with faster texture processing, improved asset loading, configurable concurrency, better caching and several additional fixes. The results so far are honestly great - uncached textures load dramatically faster and scene loading is showing clear improvements. These changes affect core VaM loading systems, though, so I need more time for stability testing and proper comparisons before releasing it. I’m not going to publish 3.0 until I’m sure it actually works reliably.
Credits:
Included - everything is set up to work together:
Assembly-CSharp.dllandSkinMeshPartDLL.dll, with the v2.0 fixes- 3x MAX_HEAP_SECTS
mono.dllpatch - NativeGCPatcher, updated to 3.1 with cold-cache startup fix
SkinMeshPartDLL.iniandboot.configsettings
Main changes:
- Fixed native morph batch sizing and null morph-bank handling.
- Restored the vanilla managed morph fallback and morph-split lookup.
- Fixed lost restore/sync arguments in exclusive-clothing processing.
- Restored missing API contracts used by precompiled plugins while keeping additional methods.
- Added all security protections introduced in VaM 1.22.0.13.
- Fixed the NativeGCPatcher cold-cache startup case that could produce
too many heap sectionswhile VaM rebuilt a large deleted cache. - NativeGCPatcher now supports a 32 GB GC trigger. I personally use 32 GB with 64 GB of RAM. If your system has less RAM, you must set a lower value.
Requirements and installation:
- BepInEx is required only if you want to use the included NativeGCPatcher.
(You can use the CPU performance patch without BepInEx, but the ZIP also includes NativeGCPatcher, which requires BepInEx 5.
For the complete package: install BepInEx 5 first, then copy all folders from the ZIP.
Without BepInEx: skip the ZIP’s BepInEx folder and copy the remaining folders normally. The CPU performance patch works independently; you simply won’t be installing the optional NativeGCPatcher.)
- Back up the original VaM files before installing.
- Extract the release directly into the main VaM folder containing
VaM.exeand replace the existing files. - Do not combine it with another patch that replaces the same
Assembly-CSharp.dll,mono.dllorSkinMeshPartDLL.dll.
The included settings were tested on my Ryzen 7 5700X3D. They are not a universal preset.
For a different CPU, please check your core and logical processor counts, then adjust and test the settings on your own system. More threads do not automatically mean better performance, and there is no single formula that gives the best settings for every CPU and scene.
I’m also working on a much bigger 3.0 update with faster texture processing, improved asset loading, configurable concurrency, better caching and several additional fixes. The results so far are honestly great - uncached textures load dramatically faster and scene loading is showing clear improvements. These changes affect core VaM loading systems, though, so I need more time for stability testing and proper comparisons before releasing it. I’m not going to publish 3.0 until I’m sure it actually works reliably.
Credits:
- turtlebackgoofy — CPU Performance Patch
- Evo — CPU Performance Patch
- Ashu27 — Anti-Stutter Patch / NativeGCPatcher
- Neverm7nd — 3x MAX_HEAP_SECTS Mono patch
Contains executable files or external scripts. Ensure you trust the creator and apply your own security measures.
