Latest Articles

Explore technical articles from our team. From compilers, AI accelerators to high frequency trading systems, we share practical experiences and deep technical knowledge gained from delivering optimized systems for the world's most demanding computing challenges for our clients.

Linear Algebra in MLIR

MLIR has a powerful Linalg dialect for linear algebra operations. Learn how to use Linalg to simplify tensor computations and optimize performance....

Affine Dialect and OpenMP

Learn how to use the affine dialect in MLIR to express loops and array accesses that are amenable to parallel execution, and how to leverage OpenMP for explicit control over parallelization....

Introduction to MLIR and Modern Compilers

This blog post provides a comprehensive introduction to MLIR and modern compiler infrastructure, exploring how these tools are revolutionizing the development of programming languages and machine lear...

Memory in MLIR

MLIR's memory architecture represents a balance between high-level abstractions and low-level efficiency. At its core, MLIR provides three primary ways to represent data: Tensors for immutable, abstra...