Data science tools have come a long way, and Project Jupyter has been foundational to that progress. But what if we could dramatically improve their performance without abandoning the Python ecosystem?
In this talk, I’ll introduce Zasper, a high-performance IDE for Jupyter notebooks that delivers:
* Up to 5× lower CPU usage
* Up to 40× lower RAM usage
* Lower latency and higher throughput
* Massive concurrency support with minimal memory overhead
Zasper achieves this by reimplementing parts of the Jupyter server stack in Go, while staying fully compatible with the Jupyter protocol. If you’ve ever hit performance bottlenecks with traditional tools, this talk is for you.
Zasper is a reimagined IDE for Jupyter notebooks, designed from the ground up with high performance and concurrency in mind. It maintains compatibility with Jupyter’s wire protocol while replacing Python-based components with lean, efficient Go implementations.
This talk will cover:
* How the Jupyter server and protocol work under the hood
* Architectural pain points in traditional Python-based implementations
* Where Go can be introduced without compromising Python workflows
* Benchmark comparisons between JupyterLab and Zasper
* Lessons learned from building a Go-based Jupyter-compatible server
By the end of the session, attendees will have a deeper understanding of the internals of Jupyter, and how combining Go and Python can unlock a new class of high-performance PyData tools—ideal for large-scale, multi-user, or production-grade notebook environments.
Personally, I recommend that this proposal be moved to the Main Track. The tool is usable outside of STEM fields, e.g., data science, and apart from the background in the Scientific Python ecosystem, there is nothing that is tying this proposal to the "FOSS in Science" devroom.