On July 19, 2024, Sabey Data Centers hosted an event in Bellevue, WA that brought together industry experts to discuss the latest trends and innovations in AI and data center infrastructure. It was an evening of delicious food, unbeatable views at Ascend Prime Steak & Sushi and, of course, great conversations.  

The program focused on enterprise-level adoption of AI applications, and what the roadmap would look like for organizations endeavoring to deploy and implement AI for their businesses. Each presenter highlighted a different stage on the “Enterprise Roadmap to AI.” Here’s a recap of the key takeaways from the event. 

The Foundation of AI: Space and Power 

Presenter:Timothy Doyle, VP of Leasing, Sabey Data Centers 

Timothy Doyle kicked off the presentations with an overview of the growing demands AI places on data center infrastructure. He emphasized AI’s exponential growth and the corresponding increase in data center power consumption. With predictions that AI will consume ten times more power by 2026 than last year, the need for expansive space and robust power solutions has never been greater.  

At the same time, data center availability across every primary and secondary market is at an all-time low. Preleasing is the only reasonable solution for organizations planning for any AI deployment.  

Doyle highlighted Sabey Data Centers' extensive and expanding footprint and their vertically integrated services, positioning them as a leader in meeting these emerging needs. To learn more about Preleasing data center space, contact Sabey Data Centers today.  

The AI Data Pipeline Powered by VAST Data 

Presenter:Chris Morgan, VP of AI Solutions, VAST Data 

Chris Morgan delved into the intricacies of AI data pipelines, presenting a full-stack solution spanning data ingestion to model inference. He outlined the critical components required to support AI operations, including data storage, retrieval and model training.  

Morgan introduced VAST Data’s innovative storage and database solutions, designed to handle the vast amounts of data generated by AI applications. He also discussed the roles of hyperscalers, SaaS providers, AI clouds and co-location facilities in fostering AI development and deployment. 

Composable Cloud Infrastructure for AI Applications 

Presenter:Kevin Cochrane, Chief Marketing Officer, Vultr 

Kevin Cochrane’s presentation focused on the principles of cloud-native and AI-native engineering. He highlighted the importance of API-first and microservices architecture in creating flexible and scalable AI solutions. Cochrane detailed the significance of continuous integration and continuous deployment (CI/CD) pipelines and containerized services in accelerating AI application development.  

He showcased Vultr’s approach to platform engineering, emphasizing the deployment of AI-native applications within a trusted private cloud environment, ensuring data security and compliance on a global scale. 

Assessing the Shift to Liquid Cooling 

Presenter:Seth Davis, Principal & Director of Mission Critical Engineering, McKinstry 

Seth Davis concluded the presentations with an analysis of the industry’s shift toward liquid cooling in data centers. He compared liquid cooling to traditional air cooling, highlighting the superior efficiency and capacity of various liquid cooling solutions, including cold plate, single-phase immersion and two-phase immersion cooling.  

Davis discussed the implementation strategies for converting existing data centers to liquid cooling and the design considerations necessary for optimizing performance. He also pointed out the advantages of built-to-suit liquid cooling data centers, including reduced building size, enhanced energy efficiency, and increased IT capacity. 

Conclusion 

The dcDNA Seattle event provided insights into the future of AI and data center infrastructure. From foundational power and space requirements to advanced data pipelines and innovative cooling solutions, the presentations underscored the rapid advancements and critical considerations in this evolving field. As AI continues to drive the demand for more sophisticated and efficient data center solutions, events like these are essential for sharing knowledge and fostering collaboration among industry leaders. We look forward to seeing how these innovations will shape the future of data centers and AI applications. 

DcDNA may be coming your way! Look out for our dcDNA Bay Area event in October and dcDNA Manhattan event in November.