Commercial PCs are going to become more powerful and more expensive as enterprises configure them for data science, analytics and model training. As a result, you can expect your work laptop to resemble workstations in terms of specifications.
Enter the PC for the "data center stack" and AI model training. Just like generative AI is reshaping data center spending, the PC may have premium configurations in the future.hether PC demand has bottomed remains an open question. PC executives note that the devices acquired in 2021 will be replaced in the years ahead and the overall market is larger than it was before the COVID-19 pandemic. PC shipments in the first quarter were down 29% from a year ago, according to IDC.
PC executives, however, are hopeful demand will start to improve in part because enterprises will need devices capable of handling generative AI models, data science, analytics and collaboration workloads.
Speaking at a recent investment conference, HP CFO Marie Myers said that the PC is going to evolve. Myers said:
"First of all, I'd say the data science stack. There's no doubt, the data scientists need a rich configuration. We feel we're well positioned in terms of driving and delivering on that opportunity. Right now, we're really building out a stack for data scientists that I'm really excited about. And that configuration, I think, is going to be available on the market. The second piece is just the PC itself. If you're going to use things like Microsoft Co-pilot, you're going to need a richer configuration inside your laptop to drive that and enable that ability.
Finally, we see an opportunity that's probably differentiated around the fact that people are going to tend to work on the edge. A lot of what we've done has been cloud. But as data moves closer and people are really worried about privacy and security, you don't want to necessarily, build a model and have all of that data exposed in the cloud. So, there is an interesting opportunity that certainly we're starting to explore around working on the edge and what that laptop configuration would look like in that area is quite unique and different."
Sam Burd, President of Dell Technologies' Client Solutions Group, had a similar take. Speaking at an investor conference, Burd said:
"We're seeing today AI models running on workstations where you can run 5 million to 500 million parameter models. And we see this trend in business spaces or an approach where businesses are going to want to do AI in their environment, on their data, on devices that they have.
Think about that workstation environment coming to more mainstream PCs, where we see the advent of offload engines and NPUs or VPUs that add AI processing capability on CPUs that will be available on the architecture in the future."
Lenovo CEO Yang Yuanqing said on Lenovo's recent earnings conference call that "PCs are still the essential productivity tool in this digital era" and demand will pick up in the second half into 2024.
Yuanqing said hybrid work and digitization will spur PC upgrades. In addition, the definition of productivity will change.
Here's where the PC is headed.
- PCs will handle select AI workloads.
- Co-pilot-powered applications will need richer configurations.
- GPUs will take on more importance.
- Collaboration including Microsoft Teams and Zoom will use more PC power.
- CIOs see PCs as critical to increasing productivity but will look for subscription models to smooth out spending.
- Role based configurations on PCs for AI and data science will become the norm.
Also see: Dell Technologies Q1 better than expected, but sales fell 20% from a year ago
HP: AI can change the role of PCs