Add dynamic memory-based asset caching and aggressive loading

- Add MemoryMonitor class for dynamic cache sizing based on available RAM
- Increase terrain load radius to 8 tiles (17x17 grid, 289 tiles)
- Scale worker threads to 75% of logical cores (no cap)
- Increase cache budget to 80% of available RAM, max file size to 50%
- Increase M2 render distance: 1200 units during taxi, 800 when >2000 instances
- Fix camera positioning during taxi flights (external follow mode)
- Add 2-second landing cooldown to prevent re-entering taxi mode on lag
- Update interval reduced to 33ms for faster streaming responsiveness

Optimized for high-memory systems while scaling gracefully to lower-end hardware.
Cache and render distances now fully utilize available VRAM on minimum spec GPUs.

include/core/memory_monitor.hpp

@@ -0,0 +1,47 @@
+#pragma once
+
+#include <cstddef>
+#include <cstdint>
+
+namespace wowee {
+namespace core {
+
+/**
+ * Monitors system memory and provides dynamic cache sizing
+ */
+class MemoryMonitor {
+public:
+    static MemoryMonitor& getInstance();
+
+    /**
+     * Initialize memory monitoring
+     */
+    void initialize();
+
+    /**
+     * Get total system RAM in bytes
+     */
+    size_t getTotalRAM() const { return totalRAM_; }
+
+    /**
+     * Get currently available RAM in bytes
+     */
+    size_t getAvailableRAM() const;
+
+    /**
+     * Get recommended cache budget (30% of available RAM)
+     */
+    size_t getRecommendedCacheBudget() const;
+
+    /**
+     * Check if system is under memory pressure
+     */
+    bool isMemoryPressure() const;
+
+private:
+    MemoryMonitor() = default;
+    size_t totalRAM_ = 0;
+};
+
+} // namespace core
+} // namespace wowee

include/game/game_handler.hpp

@@ -965,6 +965,7 @@ private:
     bool taxiActivatePending_ = false;
     float taxiActivateTimer_ = 0.0f;
     bool taxiClientActive_ = false;
+    float taxiLandingCooldown_ = 0.0f;  // Prevent re-entering taxi right after landing
     size_t taxiClientIndex_ = 0;
     std::vector<glm::vec3> taxiClientPath_;
     float taxiClientSpeed_ = 18.0f;  // Reduced from 32 to prevent loading hitches

include/pipeline/asset_manager.hpp

@@ -104,7 +104,7 @@ private:
     mutable std::mutex readMutex;
     std::map<std::string, std::shared_ptr<DBCFile>> dbcCache;
 
-    // Decompressed file cache (LRU, 1GB budget for modern RAM)
+    // Decompressed file cache (LRU, dynamic budget based on system RAM)
     struct CachedFile {
         std::vector<uint8_t> data;
         uint64_t lastAccessTime;
@@ -114,7 +114,7 @@ private:
     mutable uint64_t fileCacheAccessCounter = 0;
     mutable size_t fileCacheHits = 0;
     mutable size_t fileCacheMisses = 0;
-    static constexpr size_t FILE_CACHE_BUDGET = 1024 * 1024 * 1024;  // 1GB
+    mutable size_t fileCacheBudget = 1024 * 1024 * 1024;  // Dynamic, starts at 1GB
 
     /**
      * Normalize path for case-insensitive lookup

include/rendering/terrain_manager.hpp

@@ -272,9 +272,9 @@ private:
 
     // Streaming parameters
     bool streamingEnabled = true;
-    int loadRadius = 2;      // Load tiles within this radius (5x5 grid)
-    int unloadRadius = 3;    // Unload tiles beyond this radius
-    float updateInterval = 0.1f;  // Check streaming every 0.1 seconds
+    int loadRadius = 8;      // Load tiles within this radius (17x17 grid)
+    int unloadRadius = 12;   // Unload tiles beyond this radius
+    float updateInterval = 0.033f;  // Check streaming every 33ms (~30 fps)
     float timeSinceLastUpdate = 0.0f;
 
     // Tile size constants (WoW ADT specifications)
@@ -300,7 +300,7 @@ private:
     std::unordered_map<TileCoord, CachedTile, TileCoord::Hash> tileCache_;
     std::list<TileCoord> tileCacheLru_;
     size_t tileCacheBytes_ = 0;
-    size_t tileCacheBudgetBytes_ = 8ull * 1024 * 1024 * 1024; // 8GB for modern systems
+    size_t tileCacheBudgetBytes_ = 8ull * 1024 * 1024 * 1024; // Dynamic, set at init based on RAM
     std::mutex tileCacheMutex_;
 
     std::shared_ptr<PendingTile> getCachedTile(const TileCoord& coord);