NVIDIA’s Jensen Huang & AMD’s Lisa Su Touch Down in Taipei as Computex Showdown Looms, Showcasing Next-Gen Technologies

NVIDIA and AMD CEOs are in Taiwan ahead of Computex 2026, where both firms are expected to unveil AI- and consumer-focused product updates. NVIDIA’s Jensen Huang is set to host partner meetings and a Taiwan GTC event on June 1, while AMD’s Lisa Su highlighted deeper collaboration with TSMC and said AMD is the first to reach mass production of a 2nm HPC product, EPYC Venice. The article is event-driven and supportive of sentiment, but it contains no quantified financial results or guidance change likely to move shares materially.

Analysis

The immediate read-through is not just that two CEOs are in-market, but that Taiwan is becoming the coordination point for the next wave of AI capex. That matters because the bottleneck has shifted from chip design to supply orchestration: advanced packaging, leading-edge wafers, HBM allocation, and board-level integration are now the scarce inputs, so any guidance tone from these meetings can move the entire AI hardware complex more than the product launches themselves.

Relative winner is TSM, because every incremental competitive push from NVDA and AMD increases leverage to the foundry and advanced-node ecosystem. The more aggressive the product cycle, the more it reinforces a “winner-take-most” supply chain where customers must pre-commit capacity 6-18 months ahead; that tends to compress inventory risk for TSM while extending visibility on its premium mix. The second-order loser is INTC: even if the company is absent from the stage, faster cadence at NVDA/AMD raises the bar on process competitiveness and makes any catch-up narrative more fragile over the next 2-4 quarters.

For NVDA, the setup is a catalyst for expectations, not necessarily fundamentals, which creates asymmetry around the event window. If the company leans into consumer and AI simultaneously, the market may re-rate near-term product breadth, but the bigger swing factor is whether supply can support demand without margin dilution or delayed shipments; disappointment there would hit the stock faster than any headline enthusiasm can sustain it. AMD has a cleaner near-term relative story because every credible AI expansion helps it close the perceived gap in CPU and accelerator share, but the stock is also more vulnerable to a “show-me” reaction if launch details are light on volume commitments.

Contrarian view: consensus is likely underpricing how much of this is already in the tape. The better trade may be to own the plumbing rather than the headline makers, because event-driven upside in semis is often fleeting while supply-chain beneficiaries get a longer-duration rerating. Also, if these launches prove more consumer-facing than capacity-expanding, the market could fade the move within days and rotate back to AI infrastructure names with clearer order visibility.

Key Decisions for Investors

• Long TSM vs short INTC for the next 1-3 months; thesis is that leading-edge supply concentration and faster product cadence deepen the gap, with TSM carrying lower execution risk and better visibility.
• Buy NVDA on post-event pullbacks, not pre-event strength; use a 2-6 week horizon and target a 2:1 upside/downside only if management signals supply-secure AI demand, otherwise fade any initial pop.
• Pair trade long AMD / short INTC into and through Computex over the next 1-2 quarters; AMD has more optionality from AI + CPU share gains, while INTC remains the clearest relative beneficiary if expectations get aggressively reset lower.
• Add TSM calls or equity on any dip tied to event volatility; the risk/reward is attractive over 3-6 months because every additional AI launch cycle reinforces foundry pricing power and capacity tightness.
• Avoid chasing consumer-semiconductor beta until after the announcements; if the messaging is product-heavy rather than order-heavy, the move is likely to mean-revert within days.