DO NOT UPDATE THE MODS IN THE MODPACK, OR THE MODPACK WILL BREAK!!!
🌀 Velocifine is more than just a modpack; it is a complete reimagining of OptiFine for the modern age of Minecraft. Built on the Fabric loader, it replaces the aging code of the past with a high-performance suite of modern optimization mods, meticulously tuned to deliver the smoothest experience possible.
✨ Legacy Features, Modern Power
I've brought back everything you loved about OptiFine, but reworked from the ground up for superior performance:
🚀 Performance Without Compromise
While OptiFine was a "all-in-one" solution, Velocifine uses a modular approach. This allows us to optimize specific parts of the engine—like chunk rendering, light calculation, and thread management—independently and more aggressively.
Native Hardware Acceleration: Taking full advantage of modern CPU instructions like AVX-512 (perfect for AMD Zen 4/5 and Intel 10th+ Gen).
IMPORTANT READ!!! (For 1.21.11): You may experience minor lag spikes immediately after world creation or joining. To significantly reduce initial loading screen times, the modpack prioritizes getting you into the game faster by offloading some chunk pre-generation to the first few seconds of gameplay. Performance will stabilize completely within moments as the background engine completes its initial tasks. Just delete the mod: Ksyxis if you want to have a preload world instead of faster loading screens.
JMV ARGUMENTS (Recommended):
For Intel 12th Gen or newer & Ryzen X3D processors:
-Xms2G -Xmx4G -XX:+UseZGC -XX:+ZGenerational -XX:+AlwaysPreTouch -XX:MaxGCPauseMillis=5 -XX:ZAllocationSpikeTolerance=5 -XX:+ParallelRefProcEnabled -XX:+UnlockExperimentalVMOptions -XX:ErrorFile=./hs_err_pid%p.log
For standard processors:
-Xms2G -Xmx4G -XX:+UseZGC -XX:+ZGenerational -XX:+AlwaysPreTouch -XX:MaxGCPauseMillis=20 -XX:ZAllocationSpikeTolerance=2.0 -XX:+UnlockExperimentalVMOptions -XX:ErrorFile=./hs_err_pid%p.log
Some memory recommendations to avoid lag during gameplay (Optional):
Vanilla: -Xms2G -Xmx4G (Perfect for a lightweight, smooth experience).
Vanilla + Shaders: -Xms3G -Xmx6G (Provides the extra overhead needed for visual enhancements).
Technical Gameplay: -Xms4G -Xmx8G (Optimized for massive farms, redstone machines, and high entity counts).
Technical Gameplay + Shaders: -Xms5G -Xmx10G (The ultimate configuration for high-end performance with heavy visuals).
(In-game test to memory usage with 32 preloaded chunks with 10 chunks visible)
JRE Version (Fully recommended for 1.21.11):
Eclipse Adoptium JRE 22 x64 Bits
Texture Pack configuration for the connected textures like optifine:
Extra Performance Config (For 1.21.11):
Please adjust the threadtweak and c2me configuration files in your /config directory to better match your hardware specifications
Example for threadtweak:
Tailored for 8-Core/16-Thread CPUs (Ryzen 7 9800X3D & High-end Intel)
{
"threadCount": {
"bootstrap": 1,
"main": 14 # Max performance for 16-thread CPUs (2 threads left for OS/Background)
},
"threadPriority": {
"game": 8, # High priority for smooth visual rendering and input response "bootstrap": 1,
"main": 6, # Efficient background data processing without interrupting render "io": 4, # Balanced disk priority to prevent frame drops during autosaves "integratedServer": 8 # Ensures mod logic stays synced with the game world
}
}
Example for c2me:
Global Parallelism: Optimized for 16-thread CPUs (e.g., Ryzen 9800X3D / Intel i7-i9).
# Uses 12 threads for chunk tasks, leaving 4 for render and background apps.
globalExecutorParallelism = 12
[vanillaWorldGenOptimizations.nativeAcceleration]
# Native Acceleration: Uses AVX-512 instructions.
# Essential for AMD Zen 4/5 and Intel 10th Gen+ (with AVX-512 support).
# Significantly accelerates terrain and biome generation.
allowAVX512 = true
enabled = true
[noTickViewDistance]
# Max Concurrent Loads: Eliminates "invisible walls" during fast travel.
# Optimized for high-speed L3 Cache (AMD 3D V-Cache / Intel Smart Cache).
maxConcurrentChunkLoads = 48
[ioSystem]
# GC-Free Serializer: Reduces Java Garbage Collection overhead.
# Minimizes micro-stutters when saving world data to disk.
gcFreeChunkSerializer = true
