High-Performance WebGL Apps with LLJS and asm.js

on YouTube ↗33m

" summary

Describes how C/C++ compiled to JavaScript via Emscripten and a spec called asm.js yields near-native speed by using 32-bit integer tricks, typed arrays and no GC; shows a lljs extension that adds C-like pointers and stack allocation so vector math compiles to raw heap reads and writes. Demonstrates a WebGL cloth physics demo running ~20 ms in Firefox (~35 ms in Chrome), explains AOT compiled machine-code caching, and notes missing dynamic allocation and bugs.

AI-written summary · about

" chapters (11)

1. 0:00Start
2. 0:14James Long
3. 6:17No Garbage Collection
4. 12:53Chrome
5. 14:36Current Benchmarks
6. 16:40Low-Level Javascript
7. 18:34Constructor Function
8. 22:08Pointers
9. 25:35Dynamic Allocation
10. 27:19Cloth Demo
11. 31:15Live Editor

" related talks

• 30mNative Speed on the Web: JavaScript and asm.jsAlon Zakai · Strange Loop 2013
• 34mBack to the FUTURE! From C++ to JSAlejandro Mercado · Conf42 JavaScript 2020
• 29mWebAssembly - the Final FrontierAlbert Brand · AmsterdamJS Conference 2017
• 23mA Love Story Between JavaScript and C++Fredy Andrade · JSConfCO 2018
• 41mCreating High-Performance Web Apps with WebAssemblyKonstantin Möllers · DevDays Europe 2019
• 46mTurbocharging Client-Side Processing: Leveraging asm.jsVladimir Vukicevic · Spotlight: Advanced JS 2013
• 28mWebAssembly: Of Portability, and PerformanceDeepti Gandluri · Node+JS Interactive 2018
• 39mWebAssembly for the rest of usJan-Erik Rediger · Codemotion Amsterdam 2017