MinIO’s Post

MinIO just reached an amazing milestone, one that no storage company has even come close to - one billion Docker pulls. Thank you to our customers, partners and our community for helping us get here. Read Garima Kapoor's thoughts here: https://hubs.li/Q01mQc9X0 #opensource

Turn your Snowflake instance into a modern #datalake with Snowflake’s external table functionality and MinIO for scalable data storage and management. By treating data in MinIO as read-only tables in Snowflake, you can avoid the need for data migration. Integrating Apache Iceberg further improves your modern datalake architecture, offering improved performance through efficient data partitioning and advanced capabilities like schema evolution and time travel. Snowflake’s investment into external tables highlights the importance of flexible and efficient #data solutions, whether your data is on-prem, in the #publiccloud, or at the edge. Check out this blog post to see how these improvements can help you optimize your data strategy: https://hubs.li/Q02GN-8T0

In today’s day and age, it’s critical to have a modern, scalable and high-performance #objectstorage optimized for #AI infrastructure—which is why organizations are moving to MinIO. Transitioning from the now legacy Hitachi Content Platform (HCP) to MinIO has never been easier, thanks to our HCP-to-MinIO tool. Developed to support our customers' evolving storage needs, this tool is freely available on GitHub and greatly simplifies the migration process. https://hubs.li/Q02Fbnz60

AI is driving massive increases in storage requirements—often 4-10x. In parallel, that storage has to perform at scale—exascale to be specific. This is changing the storage landscape as enterprise IT moves away from file and block, architecting in #objectstorage instead. Grant Gross from CIO Online nails it in this well-researched article about how storage is the unsung hero of #AI deployments. It features some excellent commentary from MinIO CTO Ugur Tigli about scale, drive selection and the economics of AI in the #privatecloud (save 60% over the public cloud). Quick and insightful—you don’t want to miss this read: https://hubs.li/Q02H04Qp0

AI and ML model are as good as the #data that underpins them. No one debates this fact. But what is the next most important attribute? We will argue that accessibility and performance are next and that they are massively important from a competitive perspective. We believe this so passionately that we built the industry's top performance assessment tool, WARP. This #opensource, full-featured, S3 performance assessment software is purpose built to conduct tests between WARP clients and #objectstorage hosts—making it ideal for benchmarking your #AI data infrastructure. Go deeper here: https://hubs.li/Q02D4ZPR0

We are thrilled to share the release of MinIO DataPod, a first-of-its-kind reference architecture for building data infrastructure to support exascale #AI and large-scale #datalake workloads. As object storage has become foundational for AI infrastructure, MinIO continues to innovate—building the blueprint for storing massive volumes of structured and unstructured data and providing performance at scale. Get all the details here: https://lnkd.in/gvrzepfP

The MinIO Certification program is designed to validate your MinIO skills, boost your credibility and level up your career. Remember—there’s no #AI without #objectstorage. So why not get ahead? Get #miniocertified. https://hubs.li/Q02HwrXw0

MinIO customers validate why MinIO is the unsurpassed choice for high-performance #objectstorage. Learn more about MinIO here: https://min.io/ #data #ai

With all the buzz in the industry today regarding #LLMs with their encoders, decoders, multi-headed attention layers, and billions (soon trillions) of parameters, it’s tempting to believe that good #AI is solely a matter of model design. Unfortunately, this is not the case. Good AI requires more than a well-designed model. It also requires properly constructed training and testing #data. In this post, AI/ML SME Keith Pijanowski introduces the concept of data-centric AI, a term first coined by the folks at Snorkel AI, and talks how to use Snorkel Flow with MinIO to create a training pipeline for your AI workload. Check it out. https://hubs.li/Q02HsCry0