Lenovo introduced new servers, platforms, and services designed to support artificial intelligence (AI) inferencing at Tech World during CES 2026 in Las Vegas. The offerings expand the Lenovo Hybrid AI Advantage portfolio and focus on helping organizations run trained AI models in real time across data centers, cloud environments, and edge locations.
AI inferencing refers to using trained large language models (LLMs) to analyze new data and make decisions instantly. This is different from training models, which requires heavy computing resources over long periods. By shifting attention to inferencing, companies can turn earlier investments in AI into practical tools used in daily operations, such as retail transactions, manufacturing systems, and healthcare decision support.
Lenovo said the new servers are built to handle inferencing workloads of different sizes and deployment needs. Futurum estimates the global AI inference infrastructure market will grow from $5 billion in 2024 to $48.8 billion by 2030, showing rising demand for systems that can run AI models efficiently in production environments.
The new lineup includes three main servers. The ThinkSystem SR675i is designed for large workloads and can run full LLMs in environments such as manufacturing, healthcare, and financial services. The ThinkSystem SR650i targets data centers that need high-density GPU computing while remaining easy to deploy. The ThinkEdge SE455i is a compact server built for retail, telecommunications, and industrial settings, allowing AI processing close to where data is generated.
Lenovo said the servers work with its Neptune air and liquid cooling systems to manage power use and heat. The company also offers TruScale, a pay-as-you-go pricing model that allows organizations to scale infrastructure without large upfront costs.
Beyond hardware, Lenovo introduced updated Hybrid AI platforms that combine servers with storage, networking, software, and orchestration tools. These platforms include options built with Nutanix, Red Hat, and Canonical Ubuntu Pro, each aimed at different enterprise needs, from centralized inference environments to simpler deployments for teams starting with AI.
The company also expanded its Hybrid AI Factory Services, which include advisory, deployment, and managed services. These are intended to help organizations set up inferencing environments faster, maintain system performance, and manage infrastructure over time through Lenovo Premier Support and flexible financing options.
Lenovo highlighted the Sphere in Las Vegas as an example of its enterprise technology in use. Hundreds of Lenovo ThinkSystem servers support content creation at Sphere Studios, showing how the same infrastructure can be applied to large-scale, real-world workloads across industries.