Lower-latency OpenMP-style minimalistic scoped thread-pool designed for 'Fork-Join' parallelism in Rust and C++, avoiding memory allocations, mutexes, CAS-primitives, and false-sharing on the hot path 🍴

Fast and memory-efficient immutable collections and helper data structures

Handle duplicate Datomic transactions with ease (on-prem).

Performance testing contention queues

A sandbox to play with concurrent algorithms and data structures

Atomic Operations API

Implementing concurrent queue using 2 lock approach and non-blocking CAS approach

Software Transactional Memory based on Transaction Locking 2 (TL2) algorithm.

Basic examples of compare and swap.

This project aims at presenting different synchronization techniques

Package atomic provides low-level atomic memory primitives useful for implementing synchronization algorithms.

A repository for performing deletion of Keyframes and Mappoints that are marked as bad(Logical End of life of the Keyframes and Mappoints) using reference counting and compare and swap.

Really simple C implementation of parallel queue. Three types, lockfree, with locks and monothread queue.

High-performance, operator-class Fintech Core payment processor built with Go 1.25 Workspaces under strict PCI-DSS segregation boundaries. Features an immutable double-entry accounting Ledger, an oriented FSM state graph, and a lock-free CPU-atomic CAS TPS rate limiter.

A Log-Structured Merge Relaxed Radix Balanced Vector

Compare and swap workshops.

Monitor Synchronization in Examples

A high-performance striped resource pooling implementation

Concurrency-safe ticket booking system in Java. Lock-free CAS-based seat reservation, Postgres persistence, real-time seat map via SSE. Built to prove exactly one winner when many threads race for one seat.

A real-time stock trading engine built in Java that reliably matches Buy and Sell orders under concurrent load. By using lock-free data structures, the system avoids traditional synchronization bottlenecks, enabling higher throughput and predictable performance in multi-threaded environments. Orders are matched using price-priority logic with O(n)