"My latest industry checks indicate that MediaTek has upgraded the strategic positioning of its AI business from 'IC / ASIC design' to 'system-level design,' initially targeting the PCBA (L6) for Google's TPU and the L10 rack for Elon Musk-affiliated companies' in-house AI chips," Kuo wrote. The analyst frames the move as a calculated long-term repositioning rather than a near-term earnings catalyst, noting that it carries "immaterial impact on fundamentals over the next two years."

The gross margin ambition embedded in the new strategy is notable. TF International Securities expects MediaTek to target at least 40–50% in its system-level integration business by adopting an asset-light model in which it leads design and validation, draws on Taiwan's hardware supply chain ecosystem, and outsources manufacturing. That structure would allow MediaTek to capture the engineering and integration value of rack-level complexity without shouldering capital-intensive production costs.

Kuo sees two structural forces making this pivot timely. Increasing server rack design complexity, driven by the adoption of technologies such as co-packaged optics (CPO) and 800V high-voltage direct current (HVDC) power delivery, is raising the engineering value-add of system-level work. At the same time, refresh cycles for AI server infrastructure are beginning to resemble those of consumer electronics, creating more frequent design-win opportunities for companies with deep integration expertise.

The two initial targets, however, are not equally accessible. On Google, Kuo is candid: "Google's hardware assembly ecosystem is already well-established, so MediaTek's chances of winning L10 are slim." MediaTek's more realistic entry point with Alphabet is at the PCBA level, aiming to begin with TPU v10, codenamed Icefish, while also pursuing integration of its own CPO solution.

The Musk-affiliated opportunity is structurally different. "The AI compute currently being built out by Musk's companies relies mainly on Nvidia chips, so the assembly ecosystem for their own AI chip racks is not yet in place, and that's where MediaTek's opportunity lies," Kuo wrote. Because those companies are still in early stages of developing proprietary AI silicon at scale, the rack assembly supply chain around that hardware remains nascent, giving MediaTek a potential opening that does not exist with Google. Kuo adds, however, that "long-term success will hinge on whether MediaTek can leverage Taiwan's hardware supply chain ecosystem and its partnership with Terafab to land L10 rack orders" — and that this business "still lacks clear timeline visibility."

Overall, Kuo characterizes the strategic logic as sound: "This strategic shift aligns with industry trends and, if executed well, should help strengthen MediaTek's customer relationships and long-term competitive advantage." The intent, in his words, is "to capture new growth opportunities while mitigating potential risks."

The risk TF International flags most prominently is the potential obsolescence of MediaTek's existing ASIC design growth engine. The firm warns that growth momentum in ASIC design may begin to slow within two to three years as the Semi-COT (semiconductor chip-on-trend) business model takes hold across the industry. That looming headwind is a key reason Kuo views the system-level pivot as a necessary hedge, even if it contributes little to near-term revenue.

Shares of Alphabet are trading in Monday's session against a backdrop of sustained investor interest in AI infrastructure spending, with the company's TPU roadmap a closely watched signal of its intent to reduce dependence on third-party silicon. For MediaTek, the analyst's assessment offers a window into how Taiwan's chip design champions are repositioning for the rack-level integration layer that is fast becoming the competitive frontier in AI hardware.

The clearest near-term milestone to watch is any qualification progress on TPU v10 (Icefish), which would mark MediaTek's first concrete foothold in the system-level segment Kuo describes. On the Musk-affiliated side, timeline visibility remains limited, meaning investors should treat that opportunity as optionality rather than a near-term revenue contributor. Kuo's overarching verdict frames this as a multi-year repositioning play whose payoff, if it materializes, will be measured in customer stickiness and competitive moat rather than quarterly earnings beats.