Ethan's ConcurScale Labs

Insights into building and optimizing operating systems, concurrent and parallel algorithms, distributed systems, and scalable machine‐learning architectures


CY86→x86-64 Translator: Complete Implementation with Performance Optimizations, From Zero to Working Compiler

C++ implementation of a CY86 assembly language translator targeting x86-64 ELF executables, achieving 18% performance improvement through advanced compiler and codegen optimizations

207 min read   ·   2025

Lock-Free Queues with Advanced Memory Reclamation: A Deep Dive into Epoch-Based Reclamation and Hazard Pointers

Understanding how modern concurrent systems solve the memory reclamation problem in lock-free data structures

226 min read   ·   2025

C++ Preprocessor Tokenizer: A Complete Implementation of Translation Phases 1-3

Production-quality C++ implementation of preprocessing phases 1-3 with full UTF-8, trigraph, and universal character name support, achieving 24% performance improvement through advanced optimizations

123 min read   ·   2025

From 0.37x to 18.7x: Building a High-Performance SIMD Library with AVX-512 Speedups in Data Science, Inference, & HPC Workloads

A comprehensive technical journey through building a high-performance SIMD library, achieving extraordinary speedups through masked operations, multiple data types, and advanced CPU feature detection.

290 min read   ·   2025

Level 3 mini_malloc: A Security-Enhanced Memory Allocator with Debugging Features

Technical deep-dive into mini_malloc - a memory allocator showcasing security-enhanced design patterns and debugging infrastructure. Demonstrates arena-based concurrency, immediate coalescing, dual allocation strategies, and corruption detection mechanisms. Features complete implementation (~800 lines), comprehensive test coverage, and detailed performance analysis comparing against system malloc.

210 min read   ·   2025

Modern Concurrent Red-Black Tree Design in C++: A Practical Guide - Part 1

Sharing ideas on making a Red-Black Tree thread-safe using C++ shared_mutex and multiple reader strategies, plus visual flowcharts for insert/delete under concurrency

27 min read   ·   2025