NEW enterprise-grade servers and services aimed at accelerating real-time artificial intelligence inferencing were unveiled Jan. 6 during CES 2026, as organizations increasingly move from training large AI models to deploying them in live operational environments.
Lenovo, a global technology provider with a broad infrastructure and devices portfolio, said the new offerings are designed to support AI workloads across data centers, cloud platforms and edge locations, enabling faster analysis of live data and lower-latency decision-making.
The launch reflects a wider shift in enterprise AI adoption, where inferencing — using trained models to process new, unseen data — has become central to extracting business value from earlier investments in model training. Industry estimates cited during the announcement project rapid growth in demand for inferencing infrastructure over the rest of the decade, driven by use cases in retail, manufacturing, finance and health care.
The newly introduced systems span large-scale data center servers capable of running full large language models, as well as compact, ruggedized platforms intended for edge deployments in environments such as retail outlets, telecommunications sites and industrial facilities. The company said the systems are built with enhanced graphics processing, memory and networking capabilities to support high-throughput AI workloads.
Energy efficiency and deployment flexibility were emphasized as key design considerations, with the servers incorporating both air and liquid cooling technologies. The offerings are also available through a consumption-based pricing model that allows enterprises to scale capacity without significant upfront capital spending.
Alongside the hardware, the company expanded its hybrid AI portfolio with prevalidated platforms that integrate servers, storage, networking and software orchestration. These platforms are intended to reduce deployment complexity and help organizations operationalize AI inferencing more quickly while managing performance and cost.
New advisory, deployment and managed services for AI inferencing were also introduced, targeting enterprises at different stages of AI adoption. The services are positioned to help organizations design, deploy and operate inferencing environments while maintaining reliability and scalability as workloads grow.
The rollout highlights the growing focus on operational AI, as enterprises seek to turn trained models into tools that can act on data in real time, closer to where that data is generated.
