MediaTek’s stock has surged over 20% this week, marking its best performance in more than a decade, driven by Google’s AI chip advancements and collaborations on tensor processing units (TPUs) that position the company as a key player in AI hardware alongside Nvidia.
-
MediaTek partners with Alphabet on TPUs, emerging as a rival to Nvidia in AI applications despite challenges in smartphone sales.
-
Analysts from Morgan Stanley upgraded MediaTek’s rating to overweight, citing AI upside offsetting China market headwinds.
-
UBS raised 2027 TPU sales forecast to $4 billion, with TPUs potentially contributing 20% of operating profit by 2028, supported by shifts in AI inference demand.
MediaTek stock surges 20% on Google AI chip deals, boosting outlook amid TPU innovations. Discover analyst upgrades and revenue projections for 2026-2028. Stay informed on semiconductor trends driving tech growth.
What is Driving MediaTek’s Recent Stock Surge?
MediaTek’s stock surge stems primarily from advancements in AI technologies by its key client, Google, including new deals for tensor processing units (TPUs) and the Gemini AI model. The Taiwanese chipmaker’s shares have climbed for five straight days, achieving a weekly gain exceeding 20%, the strongest in over ten years. This momentum reflects growing investor confidence in MediaTek’s role in AI hardware, even as it navigates pressures in other segments like smartphones.
How Are MediaTek’s TPU Collaborations Impacting Its Financial Outlook?
MediaTek’s partnership with Alphabet Inc.’s division focuses on designing TPUs, specialized chips for AI workloads that are increasingly seen as competitors to Nvidia’s dominant GPUs. Despite a challenging business environment in China, as highlighted by Morgan Stanley analysts Daniel Yen and Charlie Chan, the TPU initiatives are expected to counterbalance declines in smartphone revenues. These analysts, in their November 27 report, upgraded MediaTek’s stock from equal-weight to overweight, emphasizing that AI developments provide significant relief. However, the stock remains down about 2% year-to-date, underscoring ongoing margin pressures from high R&D costs and intense competition. UBS Group AG analysts, led by Sunny Lin, have similarly boosted their projections, estimating TPU contributions to reach $4 billion in sales by 2027, up from a previous $1.8 billion forecast. By 2028, this segment could represent nearly 20% of MediaTek’s operating profit, contingent on successful execution by both Google and MediaTek. The shift in AI focus from training large language models—dominated by Nvidia’s GPUs—to inference tasks is spotlighting TPUs, with additional excitement from reports that Meta intends to deploy Google’s TPUs in its data centers starting in 2027. Macquarie Group Ltd. analysts, under Arthur Lai, favor MediaTek over other Google partners and Nvidia’s supply chain for long-term AI investments, contributing to a bullish consensus with 23 buy ratings, 10 holds, and no sells. The average price target suggests a 9% upside over the next 12 months.
Frequently Asked Questions
What Revenue Projections Do Analysts Have for MediaTek’s Cloud AI Chips?
MediaTek’s Vice Chairman and CEO Rick Tsai projects $1 billion in revenue from cloud AI chips in 2026, with the first major project generating multiple billions in 2027 and subsequent ones from 2028 onward. This aligns with a total addressable market of $50 billion for data center ASIC chips, where MediaTek targets 10-15% share in the coming years, undeterred by competitors like Nvidia’s investments.
Is MediaTek’s Collaboration with Nvidia Affected by Its Google TPU Deals?
MediaTek continues its joint work with Nvidia on the GB10 Grace Blackwell Superchip, used in the DGX Spark personal AI supercomputer launched last month, without interference from Google projects. Tsai emphasized that recent developments, including Nvidia’s $5 billion investment in Intel, do not alter MediaTek’s strong positioning in AI ecosystems, allowing diversification across major tech players for sustained growth.
Key Takeaways
- AI Momentum Boost: MediaTek’s TPU collaborations with Google have driven a 20% weekly stock gain, the best in over a decade, highlighting its pivot toward high-margin AI hardware.
- Analyst Optimism: Upgrades from firms like Morgan Stanley and raised forecasts from UBS to $4 billion in 2027 TPU sales underscore potential to offset smartphone segment challenges.
- Revenue Growth Path: CEO projections point to $1 billion from cloud AI chips in 2026, with broader market share ambitions in a $50 billion data center sector signaling long-term expansion.
Conclusion
MediaTek’s stock surge and TPU-driven AI collaborations with Google mark a pivotal shift, bolstered by analyst upgrades and projections from Morgan Stanley, UBS, and Macquarie that emphasize resilience against China headwinds and smartphone pressures. With CEO Rick Tsai forecasting substantial cloud AI chip revenues and a targeted 10-15% share in the expanding data center market, the company is well-positioned for future gains. Investors should monitor execution on these initiatives, as MediaTek’s integration into major AI ecosystems promises ongoing innovation and profitability in the evolving semiconductor landscape.
The recent performance of MediaTek’s stock follows a mixed third-quarter report in October, where revenues reached $4.64 billion (NT$142.1 billion), marking a 7.8% year-over-year increase but a 5.5% quarter-over-quarter decline. Net income dipped 0.5% year-over-year to NT$25.5 billion, while consolidated net income fell 9.3% quarter-over-quarter, and operating income decreased 24.5% quarter-over-quarter and 7% year-over-year to NT$22.18 billion. These figures reflect seasonal fluctuations and competitive dynamics, yet the AI focus provides a countervailing force for recovery. Experts from sell-side research, including those at UBS and Macquarie, continue to view MediaTek’s strategic alignments with tech giants as a foundation for exceeding market expectations, particularly as inference demands accelerate in AI applications. This positions the firm not just as a smartphone chip supplier but as an essential player in the broader AI infrastructure buildout, where TPUs and ASICs are gaining traction beyond Nvidia’s ecosystem.
