Thursday 

Room 4 

13:40 - 14:40 

(UTC+02

Talk (60 min)

Allocator-Aware Rust

“Why should we be concerned with custom memory allocators in Rust?” - Most developers underestimate how much the physical location of allocated memory can make or break performance - even in code engineered for speed in all other aspects.

Platform
Technique
Language
Rust

Just replacing the global allocator with another one is not always the best/only choice. Long-running processes can quietly grind down when they lean on a one-size-fits-all global allocator. In contrast, in many workloads, local specialized allocators are better suited - allocating memory for a well-defined subset of objects in the system (eg. arena, multi-pool, monotonic).

Let’s analyze various allocation strategies using critical memory performance indicators such as: fragmentability, allocation density, diffusion, variation, utilization, locality, and contention.
Are we allocator-aware yet, in Rust? What does it take to stabilize the allocator API, so we can unlock these new found super-powers? Let’s walk through some of these challenges.

Victor Ciura

Victor Ciura is a Principal Engineer on the Rust team in Microsoft DevDiv, building the compiler toolchain and libraries needed for the broader 🦀Oxidizer effort across the organization and Rust open-source community.

Spent the last 20+ years doing systems programming in C++ on various teams, such as: Visual C++ (DevDiv), Advanced Installer, Clang Power Tools.

He’s a regular guest at Computer Science Department of his Alma Mater, University of Craiova, where he gives student lectures & workshops on algorithms and optimization techniques, using modern C++, Rust, Haskell.

More details: https://ciura.ro & linkedin.com/in/victor-ciura
Social:
🐦 Twitter/X: @ciura_victor
🐘 Mastodon: hachyderm.io/@ciura_victor
🦋 Bluesky: @ciuravictor.bsky.social