Lenovo announces AI Express developed with NVIDIA

Lenovo says AI Express combines its validated systems with NVIDIA accelerated computing, networking and AI software in ready-to-ship solutions.[1] The company says the offer is intended to address barriers to AI adoption and help customers deploy infrastructure suited to their specifications while reducing the risk of overbuilding or costly delays.[1] Lenovo also offers services that include identifying use cases, validating return on investment through proofs of concept, and optimizing GPU environments.[1] The announcement presents the configurations as a way to simplify procurement, deployment and operations, rather than as hardware alone.[1]

Small is designed for focused inference serving tens of users, with average model sizes from 7 billion to 70 billion and throughput of at least 30 tokens per second.[1] It uses a Lenovo ThinkSystem SR650a V4 with two NVIDIA RTX 6000 PRO Blackwell Server Edition GPUs.[1] Medium targets higher-throughput inference and agentic AI for hundreds of users, at least 30 tokens per second and average models from 70 billion to 400 billion; it uses an SR675 V3 with eight of those GPUs.[1] Large is intended for generative AI serving thousands of users, with models up to one trillion parameters, and pairs an SR680a V4 with NVIDIA HGX B300.[1]

Lenovo says eligible configurations can have an order-to-ship time starting at 15 business days; it lists Medium from 20 days and Large from 25 days.[1] The clock begins only after order validation, payment or credit clearance, End User Certification, and applicable due diligence or export approvals.[1] It covers shipping, not deployment, so the listed timelines should not be read as a promise that a customer’s AI system will be operational within that period.[2] Lenovo says availability, eligibility and ordering processes vary by region.[1] Its announcement did not include pricing.[3]

Customers can add Red Hat AI Factory with NVIDIA or NVIDIA AI Enterprise, and Lenovo says Veeam Kasten can be added to protect AI applications, data, models and pipelines.[1] All three configurations support current CPU technologies from AMD and Intel.[1] Lenovo also introduced Premier Support Plus for Servers, which it says uses proactive and predictive capabilities to identify issues earlier and support continuity in mission-critical environments.[1] The company’s 2026 CIO Playbook reports that organizations moving from AI experimentation to scaled execution expect an average return of $2.79 per dollar invested, while 93% of enterprise respondents anticipate positive returns from AI.[1] Through Lenovo 360, partners can bring the validated configurations and capabilities to customers.[1]