In mid-December 2025, the IT infrastructure market received a clear signal from Lenovo that the Chinese giant intends to aggressively address the gap between the growing ambitions of AI and the outdated hardware back-up of enterprises. The company announced a major refresh of its ThinkSystem and ThinkAgile portfolios, addressing two of the most pressing concerns of today’s CIOs: insufficient storage performance for AI workloads and strategic uncertainty in the area of virtualisation.
The decision to introduce new solutions is no accident and is a direct result of hard market data. According to IDC analysts, as much as 80 per cent of storage deployed in the last five years is still based on traditional rotating disks (HDDs). In the era of generative AI, such infrastructure is becoming a bottleneck, effectively stifling innovation. Lenovo is responding to this with its new ThinkSystem DS series of disk arrays. These are all-flash systems designed for SAN environments to eliminate data latency, while offering a simplicity of deployment that is often lacking in enterprise-class solutions.
Equally important, the new offering is a response to the market turmoil around virtualisation platforms. Stuart McRae, executive director at Lenovo, directly points to the “unclear virtualisation strategy” in many organisations as a barrier to modernisation. The answer is to be found in the new release of hyperconverged infrastructure (HCI) from the ThinkAgile FX family. A key differentiator of these systems is their open architecture, allowing seamless migration between VMware and Nutanix solutions without replacing the hardware layer. For the partner channel, this is a strong sales argument, offering end customers real security against vendor lock-in and flexibility in their choice of software provider.
The portfolio is complemented by solutions targeting the Microsoft and Nvidia ecosystem. The ThinkAgile MX series, integrated with Microsoft Azure Local and equipped with NVIDIA RTX Pro 6000 GPUs, clearly positions Lenovo as an infrastructure provider for edge AI processing. And for customers who prefer a Nutanix environment, there is the ThinkAgile HX series with the Nutanix Enterprise AI suite, which is expected to reduce the time to deploy machine learning models from weeks to minutes.
Complementing the hardware offensive is the expansion of the services layer. Aware of the Gartner statistic that 63% of companies do not have adequate data management procedures for AI, Lenovo is emphasising consulting and implementation services. The whole thing is bundled with the TruScale model, which is part of the market trend away from one-off CAPEX outlays to a flexible consumption model. The December launch is Lenovo’s attempt to move forward – the company doesn’t want to be just a ‘box’ supplier, but the architect of a transformation in which hardware ceases to be a brake on business aspirations.
