Node.js is great for I/O-heavy tasks - but what happens when you throw massive computations and high memory loads at it? How far can you push it before things start breaking?
After 10 years of writing JavaScript for massive datasets, I've seen Node.js crash and burn under extreme CPU and RAM pressure - and I've learned how to stop it. In this talk, I'll share real-world lessons from optimizing Node.js for heavy computation, covering what works, what doesn't, and what you should absolutely avoid.
We'll dive into worker threads, memory management, native bindings, and practical optimizations that can stretch Node.js beyond its usual limits. If you've ever found yourself battling CPU bottlenecks or dealing with mysteriously high memory usage - this talk is for you!
What you’ll take away:
✅ How far Node.js can go before CPU and memory become a problem
✅ The biggest pitfalls that slow down heavy computations
✅ Practical optimizations to boost performance under extreme load
✅ How to use worker threads, buffers, and native modules effectively
I'm a Full Stack Team Lead at Harmonya, where I lead the development of scalable, high-performance applications with a strong focus on TypeScript, JavaScript, Node.js, and React. I joined Harmonya after playing a key role at CyberSixgill, contributing to its successful exit, and after serving in 8200 in both mandatory and professional service. Along the way, I also worked as a freelancer, delivering end-to-end projects for a variety of clients and companies. Beyond tech, I'm a musician and guitarist, and in my free time, I’m passionate about education, working with teenagers to guide and inspire them. I also speak at meetups and write articles on Web development, sharing insights on development best practices, architecture, and performance improvements.