Architecture of Mathematical Computation on Kotlin
summary
Overview of KMath (a Kotlin math library) and comparison of mathematical API designs and implementations across Python, C++, Julia, Java and Kotlin. Explains the trade-offs between convenience and performance, analyzes boxing in detail, and shows how Kotlin’s context-oriented approach enables modular math libraries and interoperability with high-performance platform libraries. Concludes with discussion of whether mathematical libraries can thrive outside C++, including Kotlin’s prospects.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 4:24Why math in programming
- 7:33Two-language problem in computation
- 12:18Numbers and operations in computer math
- 20:06Performance and boxing in JVM
- 28:15Math libraries architecture
- 35:20Algebras and operation abstraction
- 45:20Working with expressions and compilation
- 51:16Performance optimization and benchmarks
- 58:18Summary and Kotlin's future in science
related talks
45mKotlin for Science
40mGraphics Programming with Kotlin
27mWhy Kotlin is my favourite example of Functional Programming?
41mKotlin for Data Science
20mUnveiling Kotlin's Power: Complex Business Simulations & Beyond
30mWhat can Kotlin do for me?
29mThe Dangerous side of Kotlin
50mWriting Java-friendly Kotlin Code