→ Back to Home
Object Storage

Hitachi Vantara Enhances VSP One Object Storage for Accelerated AI Analytics and Data Governance

Hitachi Vantara has rolled out significant enhancements to its Virtual Storage Platform One (VSP One) Object storage, specifically targeting the burgeoning demands of artificial intelligence (AI) and analytics workloads. Key among these updates is the ability to perform analytical queries directly on open-format data, such as Iceberg, Parquet, JSON, and CSV, without the need to move data to separate analytics environments. This "in-place" analytics capability is claimed to deliver up to five times faster query performance while utilizing 80% fewer CPUs. Additionally, the platform now offers up to 5.5 times faster access to small data objects, a critical improvement for many AI applications. The company has also introduced a smaller four-node starting configuration for VSP One Object, making it more accessible for organizations to deploy. This development is significant for practitioners because it directly tackles one of the most persistent challenges in enterprise AI adoption: data fragmentation and the overhead associated with moving large datasets. By enabling direct analysis on the storage platform, Hitachi Vantara is helping to eliminate data silos and reduce the latency inherent in traditional extract, transform, load (ETL) processes. This matters to data scientists, AI engineers, and DevOps teams who are constantly seeking ways to accelerate their AI pipelines and derive insights more rapidly. The reduced CPU utilization also translates to potential cost savings and a more sustainable infrastructure footprint. These enhancements align with a broader, well-established trend in cloud, DevOps, and AI, where storage is no longer merely a passive repository but an active participant in data processing. The industry is moving towards "AI-native storage" and "smart storage" solutions that are optimized for high-throughput, low-latency operations and can integrate seamlessly with AI models and analytics engines. The increasing adoption of object storage for AI workloads, with 91% of enterprises running private AI reporting meaningful use of it, underscores this shift. Vendors across the board are focusing on improving object storage performance, scalability, and integration with AI frameworks to meet the demands of massive, unstructured datasets. The ability to process data in its native format, as seen with Hitachi Vantara's update, is a crucial step towards building more efficient and responsive AI systems. In practice, this means that organizations can expect to see a tangible reduction in the time and resources required to prepare data for AI model training and inference. Practitioners should evaluate how this capability can be integrated into their existing data lake architectures, particularly if they are already using open formats like Iceberg and Parquet. The smaller starting configuration also lowers the barrier to entry for smaller teams or those just beginning their AI journey, allowing for more agile deployments. However, it's crucial to assess the total cost of ownership, including licensing, support, and integration with other tools in their ecosystem, to ensure that the benefits outweigh any potential complexities. This move by Hitachi Vantara signifies a continued convergence of storage and compute, pushing intelligence closer to the data source and enabling a more streamlined approach to AI development and deployment.
#ai analytics#object storage#data governance#hitachi vantara#vsp one#in-place analytics
Read original source