In today’s post, we will discuss the advantages of self-describing data formats. As a case study, we will examine the popular self-describing data format HDF5. After a description of HDF5’s basic features and its data model, we will follow the development of support for modern storage architectures through history.

This post covers how to use Rust and PyO3 to optimize existing Python projects. It will also give you a basic introduction to Rust on the way.

Compilers are complex programs with complex requirements. The two most widespread C compilers, GCC and Clang/LLVM, are 10–15 million lines of code behemoths, designed to produce optimal machine code for whatever arbitrary target the user desires. In this blog post I’m going to give an overview of how the Clang/LLVM C compiler works; from processing the source code to writing native binaries.

Where performance matters, we want to make sure we know what to look for and what to optimize. For that we need to measure our code by analyzing its performance with tools either provided by the language or external ones called “profilers”. This post intends to give an overview on both of these methods while introducing some tools – like Google Benchmark and perf – and explaining their functionality.

This post is an introduction to lossless data compression in which we will explore the approaches of entropy-based/statistical as well as dictionary-based compression and explain some of the most common algorithms.