GL2GPU Live Demo
Accelerating WebGL with Zero Code Changes
Welcome to the live demonstration of GL2GPU, a dynamic translator that boosts WebGL rendering performance by converting API calls to WebGPU at JavaScript runtime. This demo allows you to directly compare the performance of standard WebGL against the accelerated GL2GPU version using popular benchmarks. For a detailed explanation of the technology, please refer to our WWW 2025 paper.
The key advantage? This performance gain is achieved with no need to rewrite source code or modify the browser, making it a powerful solution for legacy WebGL applications.
Watch the Demo Now!
🐢 Standard WebGL
🐇 Accelerated with GL2GPU
How to Use This Demo
- Use the input field below to set the number of objects for the benchmark (e.g., 100,000).
- Click a benchmark button (e.g., "Aquarium") under the "WebGL" section to see the baseline performance.
- Click the same benchmark button under the "GL2GPU" section to see the accelerated performance.
- Observe the difference in frame rates and overall smoothness, especially on mobile devices or with high object counts.
Key Features of GL2GPU
- ✅ Dynamic Translation: Automatically intercepts and translates WebGL calls to WebGPU in real time.
- 🚀 Performance Boost: Achieves significant speedups (up to 87.7% on mobile) without modifying original applications.
- 📦 Zero Modifications: Works entirely via JavaScript runtime and prototype patching. No browser or app changes required.
- 🧠 Adaptive Optimizations: Uses shader translation, render state caching, uniform batching, and render bundle reuse for maximum efficiency.