ASMPT SMT Solutions says demand for AI server boards helped drive its highest-ever first-quarter new order bookings, as electronics manufacturers invest in equipment capable of handling larger, denser, and more complex assemblies.
The company closed the first quarter of 2026 with record new order bookings, with AI infrastructure and server applications identified as major drivers. High-performance computing systems require reliable placement of oversized, irregular, and high-value components on increasingly demanding boards.
AI servers are changing production requirements for electronics manufacturing. Processor packages, accelerators, memory, power devices, connectors, and thermal interfaces are becoming larger, heavier, more valuable, and more difficult to place consistently. Miniature passives and dense board layouts still have to be handled on the same assemblies, forcing production equipment to manage extremes of component size without sacrificing speed or accuracy.
ASMPT’s SMT portfolio is being developed around that mixed-component challenge. AI and high-performance computing boards can include large ball grid arrays, complex packages, and fine-pitch components on the same board, placing additional pressure on placement precision, board handling, process stability, inspection, and traceability.
The company has separately highlighted multisize placement as a growing electronics manufacturing requirement. AI and high-performance computing are pushing designers towards combinations of large, heavy, and irregular components alongside small passives and dense interconnect. SMT lines increasingly have to balance throughput against placement assurance.
Electronics manufacturers serving AI infrastructure cannot rely only on conventional speed metrics. Placement errors, package damage, board warpage, thermal mismatch, and rework risk become more expensive as component values rise. A single failed assembly can carry significant material cost before labour, lost capacity, and delivery disruption are considered.
AI infrastructure investment is moving through the full hardware supply chain. The most visible spending is often tied to data centres, GPUs, and cloud platforms, but the manufacturing burden lands across boards, substrates, power modules, cooling assemblies, connectors, cabling, test equipment, and production automation. SMT capacity has become part of the AI infrastructure build-out because servers cannot scale without qualified, repeatable electronics production.
The same pressure is visible in wider European electronics activity. Bull and Foxconn have moved to manufacture AI and cloud systems in Europe, while semiconductor and power component suppliers are aligning product development around higher rack power, higher density, and greater thermal stress. AI demand is not confined to chip design; it is pulling on every layer of the hardware stack.
SMT equipment suppliers therefore face a demanding mix of opportunity and technical expectation. Equipment must support high-accuracy placement, flexible board formats, robust software, material-flow visibility, and integration with inspection and factory data systems. Customers want throughput, but they also need confidence that complex assemblies can be built repeatedly with traceable process conditions.
The labour dimension adds further pressure. Electronics manufacturers are being asked to scale capacity while managing skills shortages and shorter product cycles. More capable placement systems, better process software, and stronger automation can reduce manual intervention, provided the systems are properly integrated into the full production line.
Europe’s interest in AI hardware manufacturing and semiconductor resilience gives the trend added industrial weight. Building AI infrastructure domestically requires more than data-centre construction and silicon strategy. It needs board-level manufacturing, high-reliability assembly, test capability, power electronics, thermal engineering, and suppliers able to respond quickly as system architectures evolve.
ASMPT’s record bookings show how quickly AI infrastructure demand is translating into manufacturing equipment investment. The companies that build the machines behind electronics assembly are becoming part of the capacity equation for AI, high-performance computing, and the wider digital infrastructure build-out.
