Enterprise IT. Lenovo

Lenovo today added a bevy of platforms optimized to run artificial intelligence (AI) applications, including offerings from third-party partners that have now joined its alliance program, including a virtual assistant that provides customer assistance via a kiosk that has been developed in collaboration with DeepBrain and NVIDIA.

In addition, Lenovo is now offering a set of services in collaboration with NVIDIA to help organizations jumpstart their AI initiatives.

Collectively these platforms, applications and services will enable organizations to deploy AI applications across a mix of different classes of processors, says Robert Daigle, global head of AI for Lenovo.

At the core of the latest editions to the Lenovo portfolio is the Lenovo ThinkPad T14s Gen 6, the first CoPilot+ PC offering from Lenovo and the ThinkStation PX desktop PC that includes graphical processor units (GPUs) and neural processing units (NPUs). In addition, Lenovo is making available ThinkSystem servers that take advantage of Lenovo Neptune liquid cooling to reduce energy consumption of AI applications by as much as 40% when running in a data center.

Beyond providing platforms optimized for AI, the Lenovo AI Center of Excellence provides access to a set of Fast Start services based on more than 165 third-party applications running on software platforms such as the NIM microservices frameworks provided by NVIDIA, including a Smart Travel application that improves safety by detecting the presence of birds that might inadvertently strike an aircraft.

Additionally, Lenovo also plans to shortly make available a set of AI Advisory services that will provide, for example, a set of discovery tools that can be used to map and visualize an organization’s overall AI strategy.

All of these offerings are the result of a $1 billion investment that Lenovo previously pledged to make to help organizations operationalize generative AI over the next three years, says Daigle.

In general, organizations are becoming savvier about what types of platforms to optimally run AI applications using a mix of GPUs, CPUs, NPUs and other accelerators, he notes. “Customers are starting to run those price/performance calculations,” says Daigle.

It’s not exactly clear when AI might drive a way of IT infrastructure upgrades. Most organizations are still establishing proof-of-concepts (PoC) to help identify the AI use cases that might provide the best return on investment (ROI). As such, it might not be until next year that AI applications find their way into production environments.

Regardless of use case, however, it’s clear existing IT infrastructure ranging from endpoints to servers is not going to be able to optimally run AI applications. As such, organizations will need to include the cost of significant infrastructure upgrades, including any technologies required to keep data centers cool, when assessing the total cost of AI.

In the meantime, given the vested interest providers of IT infrastructure such as Lenovo have in making sure there is enough AI software available to justify investments in IT infrastructure, organizations should expect to see providers of IT platforms doing everything they can to ensure organizations can successfully deploy those applications.