The Apparent Connection between CIO's Success and Business Outcomes
To the Cloud & Beyond A Death Knell for Private Data Centers?
Mastering Partnership with a Remote Data Center
Understanding the Business First
Datascience: The Three Lessons Learnt
David Elges, Chief Information Officer, DC Government
Drowning in Data? Your Enterprise Might Be an AI Candidate
Troy Lau,Division Leader for AI, Human and Data Technologies, Draper
Building a Data-Centric Ecosystem
Michael Thieme, Senior Advisor to the Deputy Director for IT and Operations, U.S. Census Bureau
Privacy and Ethics in Data Governance
Ken Knapton, Chief Information Officer, Progrexion
Thank you for Subscribing to CIO Applications Weekly Brief
How to Apply Machine Learning in Your Data Center
AI is maturing, which means that its capabilities are expanding while its costs are decreasing. These two trends will enable data center vendors to incorporate AI/ML into a broader range of their products.
Fremont, CA: Artificial Intelligence (AI) and Machine Learning (ML) technologies are a tried and true method for data center operators to maximize uptime, optimize energy usage, detect potential risks quickly, and defend against cyber-attacks.
Major hyperscalers, for instance, have developed in-house AI to support use cases such as cooling. However, by leveraging AI-as-a-Service on cloud platforms, smaller operators can reap the benefits of AI/ML as well.
By embedding AI/ML technology in their products, data center vendors are making it easier for their customers to start using AI/ML. Specialized silicon, for instance, is designed to perform complex mathematical and computational tasks more efficiently. Because most AI use cases are very specific today, these AI chips can be trained for specific tasks such as natural language processing, pattern recognition, robotics, network security, and automation.
AI is maturing, which means that its capabilities are expanding while its costs are decreasing. These two trends will enable data center vendors to incorporate AI/ML into a broader range of their products. RISC-V and other open-source technologies, for example, are lowering the barriers to purpose-built "building blocks" that can focus on efficiency, performance, and scalability in unprecedented ways. As a result, even more adoption and use cases will emerge, including among smaller data center operators who currently believe AI/ML is too expensive to implement widely or at all.
Furthermore, AI/ML can be applied to the mechanical and electrical equipment in the data center to enable actionable insights and automation, saving the operator money. This necessitates combining traditional physics-based modeling approaches with cutting-edge machine learning (ML) techniques using data from the Internet of Things (IoT) sensors. Both ML and physics-based modeling have advantages. Combining them allows for the best of both worlds to be used to solve complex data center problems involving mechanical and electrical equipment.
With 5G and related industry 4.0 use cases, there is a significant increase in demand for 'anywhere, anytime access to applications and services like smart cities, autonomous vehicles, advanced manufacturing, AR/VR gaming, and so on. Latency can no longer be tolerated. As a result, edge data centers and multi-access edge compute (MEC) capabilities are taking center stage. With compact, low-cost, and powerful hardware in edge data centers, it is now possible to run AI/ML workloads close to the user where data is generated, resulting in real-time insights and experiences delivered by highly responsive as well as contextually aware apps.