MinIO Inc., developer of a high-performance, Kubernetes-native object store compatible with Amazon Web Services Inc.’s S3 service, today announced a version of its cloud storage service designed to support all types of training data artificial intelligence in a single infrastructure. .

Object storage is a highly scalable method for storing various types of structured and unstructured data distributed across multiple hardware devices. MinIO says more than half of Fortune 500 companies use its service.

Citing research that found the main reasons organizations are adopting object storage are to support AI initiatives and provide public cloud-like performance and scalability, MinIO said its new AIStor is specifically tailored to workloads that require feature understanding stored data.

The new service includes an S3 API called promptObject, which allows users to manipulate unstructured objects as if conversing with a large language model. PromptObject function calls can be combined with chained functions to address multiple objects at once. For example, a user can inquire about abnormalities on a stored MRI scan without going through an LLM. This allows developers to extend application capabilities without requiring domain-specific knowledge of automatic retrieval generation models or vector databases, simplifying AI application development.

A private API compatible with Hugging Face Inc.’s open-source AI model repository. it allows organizations to create their own data stores and models on a private cloud or in thin-air environments without code changes, reducing the risk of data leakage.

A redesigned console user interface for MinIO storage management supports identity and access management, information lifecycle management, load balancing, firewall, security, caching, and orchestration from a single pane of glass. The console also has a Kubernetes operator that further simplifies the management of large-scale data infrastructure across hundreds of servers and tens of thousands of drives.

Support for S3 via direct remote memory access takes advantage of high-speed Ethernet by leveraging low-latency, high-throughput RDMA capabilities to improve performance with low CPU usage.

MinIO’s survey of 656 IT executives found that an average of 70% of their cloud-native storage is in object form today, and that average will grow to 75% in two years. The survey found that object stores are the primary foundation for advanced analytics, AI model training, and data lakes/lakes. More than half of respondents plan to build a data lake on an object storage foundation in the next 12 months, and 41% use or plan to use object storage to support AI workloads.

One blog postMinIO said the survey results should shock manufacturers of network-attached storage, which is prized for its low cost but does not offer the cloud-native features needed for AI development. “SAN/NAS technologies are ill-suited for the cloud-native world, and you can’t containerize an appliance,” wrote Jonathan Symonds, the company’s chief marketing officer.

Image: SiliconANGLE/DALL-E