Наталя Хандусенко Hot News 12 November 2026, 11:00

Up to 800x Speedup: Ukrainian Developer Creates Ultrafast Corundum HFT Library for Interprocess Communication

Reddit user SIIMUCH, an Embedded Developer and Software Engineer, has unveiled his own open source project, the Corundum HFT library, which is designed for ultra-fast interprocess communication (IPC) on Linux and shows results that are hundreds of times better than standard OS solutions.

Why was a new tool needed?

The developer was faced with the task of optimizing the transmission of high-frequency telemetry — for example, for reliable and fast communication between the background monitor daemon and the graphical user interface (GUI).

As the author notes in his Reddit post, traditional approaches like sockets and regular pipes are forced to constantly copy memory through the operating system kernel and perform a large number of system calls. In high-load environments, this creates noticeable delays.

How Corundum HFT works

The library architecture is completely devoid of these shortcomings thanks to a number of technical solutions.

The root of the system is based on the shm_open segment, which is shared by both processes. The scribe constructs the message (crd_reserve) directly in shared memory, and the reader (crd_read) uses it in place without any additional copying in either direction.

Using memory barriers (memory_order_acquire / release) completely eliminates heavy mutexes on the hot path. The parties agree through one load with acquire and one write with release.

A hot path consists of several inline functions in the header. The library never calls dynamic memory allocation (even during setup).

For optional blocking (crd_wait), the process sleeps on the futex only when the reader is actually asleep, while crd_notify only costs one memory barrier.

Productivity and capabilities

Thanks to this architecture, the one-way latency is only 25–75 nanoseconds (for comparison: in conventional pipes this figure reaches ~22 microseconds, which makes the new development up to 800 times faster).

Technical features and implementation details:

the buffer unit is designed for a "one writer - one reader" scheme;
records have variable length (up to half the buffer capacity), are aligned to 8 bytes and never cross the end of the buffer (lie in a single continuous block);
allows you to prepare and publish (or read and release) multiple records at once;
the reader checks the header of each record, protecting the system from erroneous or malicious scribes. The slot of an incapacitated scribe is automatically released;
The core is written in pure C11, and for C++, a convenient wrapper in C++17 has been deployed with RAII support, in-place object construction (try_emplace<T>), and typed copy-free reading (msg.as<T>()).

The kit also includes the tools corundum-inspect (for checking buffer fullness and connected processes) and corundum-bench (for comparing performance with pipes and Unix sockets).

The developer notes that the project is fully open, has test coverage (CI for x86-64 and arm64) and is accompanied by some documentation (including a detailed justification of the order of operations at docs/DESIGN.uk.md).

More details and the project source code are available on GitHub .

While the AI is writing code, you can play Doom and keep a calendar: a Ukrainian developer has created an open-source Arcade plugin for Claude Code. It has a task tracker, team chats, a calendar, and 61 games.

Ukrainian IT guy created a toolkit for Home Assistant: how to develop custom integrations and UI cards using AI

Ukrainian developer creates Folders — a file manager for macOS in the style of Windows Explorer