Building the Open Source Alternative to AWS

· Source: The Data Exchange · Field: Technology & Digital — Cloud Computing & IT Infrastructure, Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, extended

Summary

Umur Cubukcu, co-founder of Ubicloud, discusses the concept of an "open cloud" as an open-source alternative to dominant hyperscalers like AWS, Azure, and GCP. He emphasizes that a truly open cloud requires an open-source control plane for transparency, portability, and freedom from data lock-in. Ubicloud's approach involves providing core services like VMs, Postgres, Kubernetes, and GitHub Actions runners on bare metal or existing cloud infrastructure at significantly lower costs, often a third of hyperscaler prices, by utilizing newer hardware like AMD EPYC processors and NVMe disks. The company aims to address data sovereignty concerns and high egress fees, positioning itself as a non-imperial provider that allows users to maintain control over their data and infrastructure, with over 500 customers and spinning up more than a million VMs weekly.

Key takeaway

For engineering leaders and architects evaluating cloud infrastructure, consider Ubicloud's open-source alternative to mitigate vendor lock-in and reduce operational costs. Its focus on portability, transparent control planes, and competitive pricing for core services like Postgres and Kubernetes offers a compelling option, especially for regulated industries or startups seeking significant cost savings. Explore Ubicloud's free credits to test its performance and suitability for your specific workloads, particularly for CPU-heavy AI inference or CI/CD pipelines, before committing to hyperscaler-dependent solutions.

Key insights

An open cloud offers transparency, portability, and cost efficiency by open-sourcing its control plane and avoiding data lock-in.

Principles

Method

Ubicloud installs its open-source control plane on bare metal or existing cloud infrastructure, offering virtualized instances and managed services like Postgres, Kubernetes, and GitHub Actions runners at competitive price-performance ratios.

In practice

Topics

Best for: Entrepreneur, VP of Engineering/Data, Director of AI/ML, MLOps Engineer, Data Engineer, CTO

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Data Exchange.