Glenn O’Donnell, VP, Research Director
In Forrester’s 2024 predictions, we forecast that “limited chip availability will drive common sense and taper down AI expectations.” How much tapering down will actually occur? In this episode, Vice President and Research Director Glenn O’Donnell weighs in on the current state and future impact of the high-end chip shortages.
The episode starts with some background on the current chip shortage and how it differs from those in the recent past. O’Donnell says the current situation was kicked off by an uptick in semiconductor demand during the pandemic, followed by a demand drop during the economic downturn that slowed production. Then came the generative AI boom (spurring the cloud providers to stockpile certain chips), and it was a clear recipe for the high-end chip shortage we’re seeing now. The discussion also covers the impact of geopolitical issues, including trade restrictions and incentives for chipmakers to build facilities in certain countries, on the supply of high-end chips.
Why can’t supply catch up? “There are only a few companies in the world that make these [high-end] chips, so they have limitations,” O’Donnell says in providing his outlook for chip capacity in 2024 and 2025. O’Donnell points out that while NVIDIA is the most prominent maker for AI chips, some other players including Microsoft and Google are designing chips specifically for AI algorithms.
Later in the episode, O’Donnell provides suggestions to help CIOs and tech leaders manage through the shortage. He says both enterprise tech leaders and even cloud providers may have to become more “pragmatic” about their AI plans. “Prioritize your AI ambitions,” he says. “If you could only pick three things, what three things would those be, instead of the 20 or 30 that people may be bringing you to deploy?”
The episode closes with a short discussion of the sustainability impact of these new, more powerful chips and how that could eventually impact supply as well. Be sure to stick around to hear that outlook.