Authors
Published
15 Jul 2024Form Number
LP1990PDF size
5 pages, 236 KBAbstract
AI inference is well suited for edge-related use cases, providing organizations the ability to leverage frontline data in real time. However, building the right hybrid cloud infrastructure to support AI inference at the edge presents a unique set of challenges. This article discusses key considerations in building hybrid cloud infrastructure for edge-based AI inference.
Introduction
According to a recent survey by S&P Global Market Intelligence, 77% of IT leaders plan to invest in generative AI, and 96% of that group are looking to extend AI capabilities to edge locations to capitalize on real-time data processing and decision-making.1 The intense interest in AI at edge locations is not surprising given the potential to enhance customer experiences, streamline operations, and provide competitive advantage — especially leveraging inference AI.
AI inference leverages trained AI models to make predictions or conclusions from net new data. It enables immediate action wherever data is consumed, and it’s changing the way organizations operate at edge locations. There are virtually unlimited use cases for this capability. For instance, AI inference might warn of imminent equipment failure at a remote factory. It can help medical staff monitor patients and improve the quality of healthcare decision-making. AI inference can help banks monitor financial transactions at edge locations and flag suspicious activity in real time.
While AI inference unlocks new possibilities, it also creates a unique set of hybrid cloud challenges, any of which can undermine the success of a new AI project before it gains traction. Without a robust hybrid cloud infrastructure, organizations face delays, cost overruns, and the potential for slow adoption when launching AI at the edge.
(1) S&P Global Market Intelligence, “2024 Trends in Data, AI, and Analytics,” November 2023
Key considerations for AI inference at edge and ROBO locations
There are a few things to keep top of mind when planning a hybrid cloud for AI inference at the edge:
- Scalability and simplicity are critical
Solutions with preconfigured hardware and software stacks are essential for edge and ROBO use cases. Deploying new nodes should be a plug-and-play capability. Centralized management tools are also important. Search for solutions with a single pane of glass to monitor and manage both the hardware and software elements of your cloud.
- The form factor
Scaling to ROBO locations often means putting IT infrastructure in unusual locations. In remote offices, high-end IT appliances may find a home in a broom closet, a small conference room, and even under someone’s desk. In situations like this, it helps to have appliances that are purpose-built for ROBO environments, including being smaller and easier to fit into tight spaces.
- Reliability matters even more than usual
Those edge appliances under a desk or in a broom closet at remote offices? They are expensive to get to, expensive to fix, and expensive to replace — more so than equivalent data center assets. Given the downsides of repeatedly deploying IT resources to edge locations, look for the most reliable appliances possible and best-in-class high availability features.
- Choose a partner with the reach you need
The fastest way to scale AI at the edge is with a partner who can provide a single point of support for hardware and software, wherever your company does business. Keep your IT team at their desks and focused on more strategic work while a partner enables remote locations. Seek vendors who offer capabilities including 24/7 technical support, regular maintenance schedules, and fast response times. Additionally, consider solutions that provide robust remote monitoring and management capabilities to further minimize the need for on-site interventions.
- It may help to engage a design partner too
An experienced partner who works side by side with your team to design a hybrid cloud for AI that is tailored to your needs can accelerate the project and improve outcomes. From the speed of the initial rollout to the long-term security of your edge devices, augmenting your team’s capabilities during the planning stages can make a significant impact.
- Pay close attention to bandwidth and latency
Minimize latency for real-time processing and decision-making — or fail to take advantage of the full potential of AI inference at the edge, as well as the security and compliance benefits of processing data at the edge. This is especially important for applications that require immediate data analysis and response. AI is an appropriate application to invest in best-in-class infrastructure up and down the technology stack.
AI at the edge generates large amounts of traffic between remote locations and data centers. Bandwidth management is a critical success factor for this data-hungry application.
Optimize AI inference performance with Lenovo and Nutanix
Organizations seeking to optimize hybrid cloud performance for AI inference should consider the comprehensive solutions offered through the partnership between Lenovo and Nutanix. This collaboration brings together advanced hardware, software, and global services to deliver scalable, easy-to-manage, and secure AI infrastructure.
The Lenovo ThinkAgile™ HX series, running Nutanix Cloud Platform, consolidates compute, storage, and virtualization software into plug-and-play building blocks, easily managed in scale-out clusters through a single interface to simplify fleet management of large-scale edge deployments. Maximize high availability with zero-touch deployment and uninterrupted updates, data redundancy features, and cloud backup for maximum uptime.
Scale as you grow from a single node to a multi-node cluster with near-limitless edge nodes at remote locations, including the purpose-built Lenovo ThinkAgile™ HX360 V2 Edge, in a pre-validated bundle with Nutanix software and leading open-source AI frameworks to run AI inferencing workloads. Featuring an edge- and ROBO-friendly form factor, the ThinkAgile™ HX360 V2 Edge can take advantage of Nutanix Validated Design for Enterprise Edge with AI, enabling go-live within weeks.
According to a 2024 study by ESG2, ThinkAgile HX solutions with Nutanix Cloud Platform provide up to 61% reduced TCO and up to 418% ROI.
(2) Enterprise Strategy Group, “Economic Validation: The Economic Benefits of Lenovo ThinkAgile HX Series with Nutanix Cloud Platform,” May 2024
More information
For more information on how Lenovo and Nutanix can optimize your hybrid cloud for AI at the edge, visit https://www.lenovo.com/nutanix-infrastructure.
Authors
Ritu Jain is a Senior Product Manager in Lenovo and she is currently the worldwide product manager for the Lenovo ThinkAgile HX family of Software Defined Infrastructure (SDI) systems. She brings more than 10 years of experience in SDI, Converged and Hyperconverged solutions.
Amalu Susan Santhosh is the Worldwide Technical Product Manager for Lenovo’s ThinkAgile HX and MX/SXM Series of Hyperconverged Infrastructure (HCI) solutions. Amalu is responsible for showcasing the business value and differentiation of Lenovo’s hybrid cloud solutions and contributing to the product lifecycle process.
Trademarks
Lenovo and the Lenovo logo are trademarks or registered trademarks of Lenovo in the United States, other countries, or both. A current list of Lenovo trademarks is available on the Web at https://www.lenovo.com/us/en/legal/copytrade/.
The following terms are trademarks of Lenovo in the United States, other countries, or both:
Lenovo®
ThinkAgile®
Other company, product, or service names may be trademarks or service marks of others.
Configure and Buy
Full Change History
Course Detail
Employees Only Content
The content in this document with a is only visible to employees who are logged in. Logon using your Lenovo ITcode and password via Lenovo single-signon (SSO).
The owner of the document has determined that this content is classified as Lenovo Internal and should not be normally be made available to people who are not employees or contractors. This includes partners, customers, and competitors. The reasons may vary and you should reach out to the authors of the document for clarification, if needed. Be cautious about sharing this content with others as it may contain sensitive information.
Any visitor to the Lenovo Press web site who is not logged on will not be able to see this employee-only content. This content is excluded from search engine indexes and will not appear in any search results.
For all users, including logged-in employees, this employee-only content does not appear in the PDF version of this document.
This functionality is cookie based. The web site will normally remember your login state between browser sessions, however, if you clear cookies at the end of a session or work in an Incognito/Private browser window, then you will need to log in each time.
If you have any questions about this feature of the Lenovo Press web, please email David Watts at dwatts@lenovo.com.