Susquehanna: Intel, AMD Prioritize Server CPU Supply, Older Chips Flow into New PCs
Miles Bennett
A Susquehanna report finds Intel and AMD are routing their latest CPUs to servers first, pushing more older-generation processors into 2026 PC builds — a sign AI PC adoption may be slower than expected.
What happened?
Susquehanna's research shows more older Intel and AMD CPUs are appearing in newly assembled PCs in 2026.
The cause is not excess inventory — both companies are deliberately prioritizing their newest chips for the server market.
This means → the PC channel is getting last-generation silicon, not the latest product.
Why servers first?
Servers and data centers are where AI compute demand is strongest — higher ASPs, fatter margins.
When fab capacity is constrained, shipping to the highest-margin customer first is the rational call.
In plain terms = the same wafer sells for more to a cloud provider than to a PC OEM, so cloud gets served first.
What does this mean for AI PCs?
Intel's AI PC market share is already directly affected by this allocation priority.
Older CPUs in new machines likely means those PCs lack on-device AI inference capability.
This means → AI PC penetration figures may overstate reality — unit shipments are rising, but the share that can actually run local AI workloads is smaller than the headline suggests.
Content is for reference only, not financial advice.