Nvidia’s Computex Keynote Sparks a Broad AI Chip Rally: IPG Photonics, MACOM, Amkor Lead Gains

Shares of several semiconductor and AI infrastructure companies surged in Monday’s afternoon session after Nvidia CEO Jensen Huang delivered a keynote at Computex Taipei that reset expectations for the scale and duration of the AI chip cycle. The message was clear: the buildout is accelerating, and it is broader than many had anticipated.

Huang confirmed that Vera Rubin, Nvidia’s next-generation architecture after Blackwell, has entered full production. The chip promises a 10x reduction in inference token cost and requires four times fewer GPUs to train comparable models — a leap that directly addresses the cost and power constraints that have weighed on data-center operators. Thousands of Nvidia engineers worked on the platform, and system builders including Dell, HPE, Lenovo, Supermicro, and IBM are already in full-scale production. The ripple effect across the AI supply chain is immediate: each of those names will need more servers, more memory, more optical interconnects, and more chipmaking equipment.

The second major announcement carried a different weight. Huang unveiled RTX Spark, an Arm-based AI PC chip co-developed with MediaTek, and said Nvidia and Microsoft are planning to “reinvent the PC.” The chip integrates a Blackwell GPU and a Grace CPU on a single package with 128GB of unified memory, capable of running 120-billion-parameter AI models locally without a cloud connection. It is set to launch later this year on Windows PCs from Microsoft, Dell, HP, ASUS, Lenovo, and MSI. This move not only expands Nvidia’s addressable market beyond the data center but also signals a fundamental shift in how personal computing could handle AI workloads.

While the market often overreacts to headline news, the moves in individual stocks reflected a reassessment of the entire AI infrastructure complex. MACOM shares were among the most volatile, with 24 moves greater than 5% over the past year. Monday’s gain, while notable, did not fundamentally change the market’s view of the company’s long-term prospects, but it did underscore how tightly the chip ecosystem is tied to Nvidia’s roadmap. A previous sharp move occurred 13 days ago when the stock gained 4% as traders positioned ahead of Nvidia’s fiscal first-quarter earnings.

Analysts noted that Nvidia had accounted for a disproportionate share of the S&P 500’s gains leading into the announcement, making the keynote a binary catalyst for networking ASICs, connectivity silicon, memory, and packaging stocks alike. All these segments tend to move together on Nvidia read-throughs.

Supporting the bullish sentiment was a sharp drop in oil prices. Benchmark Brent crude fell 5.21% during the session, lowering expectations for near-term inflation and giving the Federal Reserve more room to cut rates. Because semiconductor stocks carry some of the longest-dated cash flows in the equity market — with most of their valuation tied to 2027–2030 AI capital expenditure — even a small rate compression can produce an outsized move. A 5% decline in oil can thus power 8% gains in chip stocks on the same day.

MACOM has surged 117% since the start of the year and, at $379.49 per share, trades near its 52-week high of $409.68 from May 2026. An investor who bought $1,000 of MACOM shares five years ago would now hold an investment worth roughly $6,524.

Also worth watching: Nvidia’s chips cost six figures, but the specialized connectors, cables, and thermal sensors that make them work often cost even more. One 90-year-old company has built a near-monopoly on this infrastructure — and with the AI boom just getting underway, its stock remains under the radar. Claim the stock ticker here for free.