Morgan Stanley: Apple is quietly hoarding TSMC’s capacity, possibly for self-developed AI server chips

According to Chasing Wind Trading Desk, Morgan Stanley released a research report on April 10, which tracked deep supply chain capacity changes to reveal Apple’s massive investment in private cloud computing (PCC).

It shows Apple's strategic focus is shifting from relying on third-party cloud services to large-scale internal heavy asset capital expenditures, providing a critical anchor for assessing Apple’s subsequent cost control and profit trajectory.

Huge wafer orders expose AI ambitions

Apple is quietly buying out TSMC’s advanced packaging capacity.

Apple’s SoIC (System on Integrated Chips) orders at TSMC will reach 36,000 wafers in 2026, and surge to 60,000 wafers in 2027.

This number is highly unusual: for context, AMD—the current largest SoIC customer—only requires 42,000 wafers in 2026.

According to IDC data, Apple’s annual shipments of high-end Macs are extremely low, with actual SoIC consumption at most no more than 1,600 wafers. Clearly, such astronomical wafer orders are certainly not for selling computers.

Targeting “Baltra” self-developed compute servers

Since conventional consumer chips are ruled out, where exactly are these expensive advanced capacities going?

Morgan Stanley’s inference points directly to Apple’s secret internal project — a self-developed AI server chip (internal code name “Baltra”) specifically built for private cloud computing.

“Based on the absolute scale of Apple’s orders… we believe most SoIC capacity is being reserved for Apple’s private cloud computing (PCC) 3nm AI ASIC (‘Baltra’).

This will replace the currently used M series Ultra processors to achieve better AI inference performance and efficiency.”

Shift in capital expenditure and core suspense

This aggressive hardware bet not only validates previous rumors about Apple’s long-term collaboration with Broadcom, but also sends a clear signal to Wall Street: Apple is highly confident in an explosion of Apple Intelligence inference demand.

Most AI workloads will be locked to Apple’s own chip-based servers, shifting its AI infrastructure from traditional operating expenditure (Opex) to longer-term capital expenditure (Capex).

However, behind this grand story of vertical integration, Morgan Stanley maintains a sober observation: Despite a clear strategic intention to control costs, the actual performance and efficiency of Apple’s self-developed AI ASIC under large-scale deployment remains the biggest uncertainty in this bet on computational power.

Risk warning and disclaimerThe market carries risks; investment needs caution. This article does not constitute personal investment advice, nor does it take into account the unique investment objectives, financial situations, or needs of individual users. Users should consider whether any opinion, viewpoint, or conclusion here is suitable for their particular conditions. Investments made based on this are at your own responsibility.