Assessing the Future for AI PCs – What IT Execs are Saying

Over the past couple of years, AI has infiltrated virtually every aspect of digital infrastructure, with an initial focus on storage and compute servers. Now, as organizations move from proofs of concept to deployment of AI projects, interest is growing in the potential that the AI PC holds. Attendees of recent CIO Roundtables have provided their insights into what to expect from this new class of device.

First, though, what is an AI PC? These devices are PCs that incorporate modern technologies that allow them to run AI models and efficiently perform AI-based workloads. Like any PC, AI PCs have a CPU and a GPU, but they also typically have a neural processing unit (NPU) and far more RAM and much larger storage than an ordinary PC. Naturally, they don’t come cheap; these more powerful devices are expected to cost 10% to 15% more than non-AI PCs.

From comments made during recent CIO Roundtables, it is clear that adoption will be slow. That isn’t surprising, given that most current AI efforts are not PC-centric. Most CIO Roundtable participants also still have questions about the use cases for AI PCs and how they will deliver greater business benefit. Some are concerned as well that the software that will make it possible to take advantage of the NPU is still lacking. The IT leaders said they are sure the apps are coming, but they are in no rush to buy the new PCs before they arrive and are ready for evaluation.

None of this means that the IT leaders have a negative perception of the AI PC. It is simply a case of moving cautiously in an area that is still developing. Meanwhile, many of the Roundtable participants have already identified job functions and groups that will best be able to leverage AI PCs when they first arrive. The takeaway is that the future for the AI PC is quite bright.

Roundtable discussions about potential use cases were fruitful. Not surprisingly, most participants see the developers of generative AI apps and predictive AI apps as likely candidates. Development of those sorts of apps will require far greater performance than is now available in legacy designs to effectively build code and to test and train models. Another use case that was frequently mentioned is image- and video-related content creation, modification, and generation. Data analysts and data scientists are also widely expected to benefit from AI PCs, and several industry-specific use cases were mentioned as well.

Although such demanding use cases will be the early focus for AI PC deployment, broader use is sure to come as tools such as Copilot and Q become part of the regular application workload. As with AI trends overall, IT teams that are looking at the AI PC are seeking the small early wins that will prove the clear business benefits of the technology. From there, they will proceed step by step; no one is going to try to boil the ocean by deploying AI PCs across the board. Nonetheless, AI PCs will clearly be a key component of AI infrastructure in the future.