Digital Realty and NVIDIA Examine the Next Steps for AI
The rapidly-evolving digital economy is forcing enterprises to address more questions and challenges regarding their IT infrastructure. Artificial intelligence is no different. The way we have to think about it has evolved as well. Here's what IT departments need to consider as they optimize AI for the digital economy:
Artificial intelligence requires leadership and support for infrastructure and software
Providing a hosted AI solution requires a new data-centric infrastructure architecture that puts data at the center of the architecture to localize data aggregation, staging, analytics, streaming, and management at global points of business presence.
Additionally, architecture design must address data gravity challenges when deploying large-scale data analytics and AI solutions. Data gravity inhibits enterprise workflow performance, raises security concerns, and increases costs, all complicated by regulatory requirements and other constraints. Platform support, storage optimization, security, protection, and cost-effective scalability are critical elements of success needed when enterprises bring AI, analytics platforms and applications to the data.
Cost-effective scalability means evaluating technology while mitigating financial risk.
According to Gartner, by the end of 2024, 75% of enterprises will shift from piloting to operationalizing AI, driving a 5X increase in streaming data and analytics infrastructures.
Dave McCrory, Vice President of Growth at Digital Realty, explained that “Data gravity is growing at an incredible rate, and the data creation lifecycle is experiencing growth in virtually every area.” For example, according to 451 Research, by 2022, 87% of IT leaders will maintain local copies of customer and transaction data for compliance1. This macro trend of data localization amplifies the effects of data gravity, and the result increases the number of locations where data is aggregated. Enterprises cannot expect optimal results from AI if their AI compute is not located in close proximity to their data.
We recently launched the first-of-its-kind Data Gravity Index™ (DGx), an inaugural report available for download on the Digital Realty Data Gravity Insights Hub, which provides insights to unlock new capabilities to manage the effects of data gravity. McCrory, who coined the term data gravity ten years ago, has been instrumental in leading efforts to develop the index which: is insights-led, has a patent-pending formula and includes tools to help enterprises better grasp the intensity of data gravity across 21 metros globally.
Managing the effects of data gravity to deploy AI at scale
Technologies like AI present an opportunity for enterprises to evaluate the impact of data gravity on a business’s IT infrastructure.
Current architectures are constrained and backhaul architectures cannot address data gravity challenges that impede enterprises from properly leveraging their AI infrastructure. The architecture needs to be inverted to a data-centric architecture bringing users, devices, networks and clouds to the data at global points of presence in neutral, multi-tenant datacenters.
We recently announced the industry’s first Data Hub solution featuring NVIDIA DGX A100 systems on PlatformDIGITAL®. This new AI-ready solution improves the ability to train, test, and deploy AI at scale by providing a proven infrastructure and platform optimized for enterprise machine learning initiatives. The combination of Digital Realty and NVIDIA DGX A100 systems brings AI computing, data adjacency and orchestration together uniquely and powerfully. It improves the performance and data compliance control necessary to support exploding volume, variability, and velocity of data creation, processing, and storage to accommodate digital business.
Digital Reality and NVIDIA experts share their knowledge of AI strategy
We recently took a deep dive into this strategy as a sponsor and presenter at the NVIDIA GPU Technology Conference (GTC), held October 5-9. GTC provided Digital Realty and Interxion: A Digital Realty Company, an opportunity to explain how PlatformDIGITAL® and NVIDIA DGX A100 powers AI in the datacenter.
Leaders from Digital Realty, NVIDIA, and Interxion spoke about how organizations see scaling AI infrastructure as their primary roadblock to meeting business needs. We also provided a full explanation of data gravity and how breaking down barriers to leveraging AI infrastructure is critical to the success of the modern enterprise.
Presenter Wes Jensen, Director of Technology Innovation and Strategy at Digital Realty, said, “GTC is a perfect opportunity for Digital Realty and Interxion to communicate to a global audience how PlatformDIGITAL is the single global data center platform to solve global coverage, capacity, and connectivity needs - bringing users, things, networks, and clouds to the data while removing data gravity barriers and scaling digital business.”
We are pleased to have partnered with NVIDIA to offer an opportunity to try the world's most powerful enterprise AI systems coupled with infrastructure, and AI orchestration software powered by our partners Core Scientific and SCAN. The Data Hub solution featuring NVIDIA DGX A100 systems gives enterprises the ability to accelerate their transformation on PlatformDIGITAL®.
Our test drive sessions are available, no matter where your data sits, for your organization to evaluate the technology with your own data scientists.
Test Drive with Core Scientific: https://www.interxion.com/products/ai_hpc_services/core-scientific-partnership
Test Drive with SCAN: https://www.interxion.com/products/ai_hpc_services/scan_partnership
To learn more about managing the effects of data gravity to deploy AI at scale, contact us.
 *451 Research, Infrastructure Imperative – IT Leader Survey Nov 2019