Omdia: AI wave brings opportunities in data storage and infrastructure, with the storage market expected to continue strong recovery by 2025.

Omdia states that the next wave of artificial intelligence (AI) will focus on AI agent deployment and private data enhancement training, which will lead to an increase in data center storage spending. Because these technologies are crucial for AI data pipelines, a significant opportunity is emerging in the data infrastructure space for major storage suppliers (cloud and enterprise). By 2023, storage spending growth is expected to slow down as cloud service providers significantly shift a larger portion of their data center equipment budgets towards GPU-based servers, gradually moving away from general servers and storage. In 2024, storage growth will show a strong recovery, expected to continue in 2025 and then stabilize until 2029. It is forecasted that by 2029, storage device shipments will grow at a CAGR of 12.5%. However, the traditional Storage Area Network (SAN) and Network Attached Storage (NAS) sub-segments have matured, and it is expected that these areas will experience slow growth or even negative growth. Key information about storage and data infrastructure: Global Distributed File System (GDFS) storage. GDFS improves data accessibility by providing a single view and data management for data stored in remote and multi-cloud locations (whether in local, cloud, or hybrid environments). AI data weaving for distributed data. As AI becomes successful, enterprises are seeking comprehensive data toolsets to help manage different data sources, integrate these data streams more effectively, and obtain carefully curated datasets for training. Resurgence of dynamic archiving and tape. AI and Machine Learning (ML) are driving data generation, and this data needs storage. While many infrequently accessed data can be sent to cloud archive services, local dynamic archiving storage is becoming a viable alternative. Specialized AI integrated storage platforms. All-in-one integrated platforms that combine data pipeline management, clusters, AI computation libraries, and AI-optimized storage are gaining popularity. Such platforms are particularly valuable for edge inference and enterprise enhancement training.