The analyst team at Goldman Sachs has downwardly revised their projections for rack-scale AI server shipments in 2025 and 2026, decreasing the estimates from 31,000 units to 19,000 units and from 66,000 units to 57,000 units, respectively. This revision stems from factors such as product transition phases, heightened production complexity, and increased uncertainty in demand. Goldman Sachs has further adjusted the target prices for companies involved in Taiwan's ODM and thermal cooling supply chains, with cuts ranging from 7% to 21%. While investments in AI infrastructure continue to serve as a pivotal growth engine for the technology sector, Goldman Sachs anticipates a more tempered pace of expansion moving forward.
