Grant Caley, the UK and Ireland solutions director for NetApp, discusses artificial intelligence (AI) and data storage in a recent podcast. He emphasizes the importance of storage scalability, performance, and hybrid cloud solutions for AI workloads. Caley highlights the need for access to hyperscalers, such as AWS, Azure, and Google Cloud, and the ability to efficiently move and manage large datasets for AI development and inferencing.
In terms of storage requirements for AI workloads, Caley explains that traditional enterprise workloads differ significantly from AI workloads which can rapidly scale to multi-petabyte installations across various cloud and on-premise environments. AI workloads involve complex data wrangling, movement, and preparation, and require high-performance storage solutions that can scale from gigabytes to petabytes to meet evolving demands. Additionally, security, data mobility, performance, and integration with various frameworks and platforms like Kubernetes, Python, and Kafka are essential for effectively managing AI data.
Caley also emphasizes the importance of a standardized data layer that can seamlessly integrate with different AI tools and frameworks, enabling customers to reduce complexity and improve outcomes. The focus on data security, zero-trust administration, and integration with tools like TensorFlow, PyTorch, and Kubernetes is crucial to ensuring successful AI deployments across different environments. By providing flexible and secure storage solutions that align with AI requirements, companies can optimize their AI workflows and maximize the value of their data assets.