Unity • Rendering • Frame Pacing • XR Optimization

I build deterministic performance systems for real-time applications — focusing on stable frame delivery, GPU/CPU bottleneck isolation, and reproducible profiling workflows.

• Diagnose CPU vs GPU bottlenecks using profiler-backed experiments
• Design controlled stress environments to reproduce performance issues
• Optimize rendering pipelines for mobile and XR constraints
• Maintain stable frame budgets (11ms / 16.6ms) under load
• Build telemetry-driven systems to monitor runtime health

The goal is simple: consistent frame delivery under real-world stress — not just high FPS in ideal conditions.

Performance problems are not random. They are measurable, reproducible, and isolatable.

My workflow:

• Establish baseline → apply controlled stress → measure delta
• Separate CPU vs GPU cost using profiler evidence
• Eliminate variance before applying optimizations
• Validate improvements with repeatable scenarios

If it cannot be reproduced, it cannot be fixed reliably.

Deterministic stress framework for analyzing rendering cost in XR.

Focus areas:

• Stereo rendering overhead
• Overdraw amplification (200+ transparent layers)
• MSAA cost scaling
• Fragment workload vs submission cost

Key finding:

• Single rendering factor produced +7ms GPU cost under stereo load while CPU remained stable → clear GPU-bound scenario

🔗 https://github.com/JamesDeRaja/XRPerformanceLab

Performance breakdown of Unity UI masking under layered transparency.

Focus areas:

• Mask pass cost in UI pipelines
• Overdraw amplification in stacked elements
• Canvas rebuild CPU impact
• Tradeoffs between visual fidelity and performance

Result:

• Clear identification of GPU-heavy UI scenarios caused by layered transparency and masking

🔗 https://github.com/JamesDeRaja/SoftMaskPro-Performance-Study

Comparison of update architectures under scale.

Variants:

• A — Per-object Update (baseline)
• B — Prefab-based Update
• C — Central Manager
• D — ECS (DOTS + Burst)

Focus areas:

• CPU scaling to 10,000+ entities
• GC elimination
• Update cost distribution
• Deterministic simulation control

Outcome:

• Measurable reduction in CPU overhead and improved scalability with ECS-based systems

🔗 https://github.com/JamesDeRaja/ECS-Performance-Improvement

Performance is a systems problem.

• Define budgets in milliseconds, not FPS
• Measure before optimizing
• Remove non-determinism
• Validate under stress, not ideal conditions

Architecture drives performance more than micro-optimizations.

• 13+ years in real-time systems and Unity development
• Worked with publishers: Voodoo, Lion Studios, Supersonic
• Built and optimized systems for hundreds of live game features
• Improved retention and performance metrics through optimization-driven design

🌐 https://james.alphaden.club 🔗 https://www.linkedin.com/in/james-de-raja/