Server makers saw an opening in public cloud computing, realizing that physical servers fill a crucial gap. Over time, IT leaders have accepted that some workloads belong on-premise, while others might operate in a hybrid model, and some work entirely in the cloud.
Right now, AI inference is catching the eye of server providers. They’re tackling issues like data loss, data sovereignty, and latency, especially when processing data from edge devices and the internet of things (IoT). Dell Technologies has rolled out an update to its Dell NativeEdge software, making it easier for businesses to deploy and scale AI at the edge. This platform allows for large-scale device onboarding, remote management, and orchestrating applications across multiple clouds. Dell boasts that NativeEdge helps businesses stay operational even during network failures, with features like virtual machine migration and automatic failover for compute and storage.
Take Nature Fresh Farms, for example. They manage over 1,000 IoT-enabled facilities using NativeEdge. Keith Bradley, their vice-president of information technology, shared how it helps them monitor real-time conditions for their produce and gain insights into packaging operations.
At the KubeCon North America 2024 conference, Nutanix announced its support for hybrid and multi-cloud AI through the Nutanix Enterprise AI (NAI) platform. This platform can run on any Kubernetes system—whether at the edge, in core data centers, or in the public cloud. Nutanix highlighted that NAI creates a consistent operating model for AI workloads, letting companies set up and scale inference endpoints for large language models in mere minutes.
HPE is also stepping up. During its AI event, CEO Antony Neri noted that many enterprise customers need small language models. They usually pick a large model that fits their needs and fine-tune it with specific data. He pointed out that many of these workloads operate on-premise due to concerns about data sovereignty and security. In September, HPE announced a partnership with Nvidia, creating what Neri calls a “full turnkey private cloud stack.” In just three clicks, customers can deploy HPE’s private cloud AI integrated with Nvidia’s technology.
Lenovo introduced its Hybrid AI Advantage with Nvidia during its Tech World event, which combines robust AI capabilities designed for reliability. This package includes customizable AI solutions aimed at helping businesses see a return on investment from AI.
Public cloud platforms like AWS, Azure, and Google all provide powerful environments for generative AI, machine learning, and inference workloads. They offer specific tools to support AI inference on IoT and edge devices, such as Amazon’s SageMaker Edge Agent and Microsoft’s Azure IoT hub.
In response to the cloud AI challenge, traditional server companies are seeing new opportunities. IT departments will continue to invest in on-premise solutions, and edge AI is becoming a focal point. Additionally, the availability of blueprints and templates is influencing decisions. Analyst Gartner points out that while public cloud providers excel at showcasing AI capabilities, they often fall short in helping organizations reach their specific AI goals.
During the Gartner Symposium, Daryl Plummer, a chief research analyst, expressed concern that tech providers focus too much on their own advancements without guiding customers toward achieving their AI objectives. He criticized major players like Microsoft and Google for highlighting possibilities without providing a clear path for practical applications. The real gap lies in domain expertise and tailored IT products and services, an area where Dell, HPE, and Lenovo are likely to strengthen partnerships with IT consulting firms.