AMD Boosts AI Hardware: Partners with AT&S to Expand Malaysia Base, Seizing Next Computing Competition
Keywords: AMD, AT&S, Kulim Malaysia, AI Arms Race, Data Centers, Advanced Packaging, PCB, Helios Platform
Introduction
As global AI infrastructure investment heats up, the chip supply chain enters a new high-intensity expansion cycle. Recently, AMD's moves in AI hardware layout again drew market attention: the company not only accelerates its core computing product roadmap but also strengthens upstream supply chain capabilities through partnership with Austrian PCB manufacturer AT&S. These actions show AMD is no longer focusing solely on single-chip performance competition, but aiming for a more complete AI infrastructure ecosystem, vying for greater market share in the data center era.
Supply Chain Boost: AT&S Malaysia Project Supported by AMD
Reports say AT&S will further expand capacity at its Kulim, Malaysia factory, with investment up to EUR 2 billion, funded by AMD and other customers. The expansion includes structural modifications to existing Plant 2 and construction of a new manufacturing base for IC substrates and advanced PCBs. For AMD, this is not just a simple supply chain cooperation, but a forward-looking layout for future high-end AI chip demand.
In AI server and high-performance computing scenarios, the importance of substrates and PCBs is rising. As chip power, bandwidth, and packaging complexity increase, traditional supply chains can no longer meet delivery requirements for new GPUs and accelerators. By locking in key capacity early, AMD reduces future delivery risks and strengthens supply chain resilience against rivals like Nvidia.
Financial Returns and Industrial Synergy
AT&S expects related capacity investments to contribute tens of millions of euros positively in FY2026-2027. The company also raised its new fiscal year outlook, showing rapid release of high-end electronics manufacturing demand. The market reacted strongly, with AT&S shares surging to record highs, indicating clearer capital market consensus on AI infrastructure supply chain growth potential.
From AMD's perspective, such investments increase short-term capital expenditure pressure but help stabilize supply, improve bargaining power, and ensure core customer orders are delivered on time. For a chip maker competing for AI data center share, hardware performance is just the threshold; timely, stable, large-scale delivery is key to commercialization success.
Data Center Business: AMD's Core Growth Engine
AMD's recent growth logic is clear: data center business is driving revenue and profit expansion. According to latest earnings, AMD's Q1 total revenue reached $10.3 billion, gross margin rose to 53%, with data center revenue hitting $5.8 billion, up 57% YoY. AI and cloud-related demand is no longer "conceptual expectations" but a real performance contributor.
Against this backdrop, AMD's continued AI infrastructure investment is unsurprising. Chair and CEO Lisa Su earlier said customer demand for the Helios platform is strong; as AMD's first rack-scale full-stack AI infrastructure solution, it is progressing chips, software, and system development as planned. MI450 series GPUs have begun sampling to key customers, with the platform expected to enter mass production and shipment in H2 this year.
Conclusion
From partnering with AT&S to expand capacity, accelerating optical interconnect and laser chip supply chain layout, to pushing Helios platform deployment, AMD is competing in AI industry with a more systematic approach. The future AI battlefield will no longer be just a direct confrontation of computing chips, but a full-chain competition covering packaging, substrates, PCB, optical interconnects, and software ecosystem. AMD's move to boost Malaysia capacity is both a response to current demand explosion and an early bet on AI infrastructure growth for years to come. In this escalating AI arms race, whoever builds a stable, efficient, and scalable supply chain faster will gain an edge in the next phase of competition.
