30% selloff in intel stock following an inaccurate 3q 2026 guidance miss. This large dip is indicative of the market's unease with the mismatch between management's forecasts and the company's actual performance. As a result, there's increased uncertainty about the short-term viability and execution prowess of Intel in the eyes of both purchasers and long-term holders.
Intel shifts 18A capacity to Intel 3 for Xeon 6 servers after missing revenue targets
The specific miss is in the Client Computing & Physical AI group. They missed by a 9% decrease in sequential revenue (to $8.1B). Intel's CFO, David Zinsner, had previously forecasted a flatish revenue, but that didn't happen. It shows that Intel has some serious hurdles to climb over in sustaining growth in the core hardware groups.
In an effort to fulfill near-term demand, Intel is shifting their advanced process (18A) manufacturing capacity at their Oregon factory. Specifically, they're pivoting that capacity to accommodate older process node (Intel 3) in order to fulfill demand for their Xeon 6 (Granite Rapids) servers. This shift means that some of the 18A production will be 'cannibalized' to fill out the older process node. This limits the full potential rampup of the new 18A node.
Intel forecasts more than a 20% decrease in PC-related revenue in 4Q due to memory and substrate dislocations. However, they still expect the Total Addressable Market (TAM) for Server CPU's to grow 5x to $400B. These are mixed signals for Intel's hardware business and there seems to be a period of transition.
In the future, Intel expects to phase in Nova Lake CPU's. Initially, they will phase in single compute tile SKUs. Intel's plan is to then phase in dual compute tile SKUs later in 2027. This shows a gradual approach to new architecture (nova lake CPUs).
Cannibalizing 18A for Xeon 6 seems like a desperate move. If they cannot hit their revenue targets for the client group how do they fund the advanced node ramp. This feels like short term thinking over long term viability.