Tech never stops moving. Just the other day, Google published a study suggesting that future quantum computers could crack the elliptic curve cryptography behind cryptocurrencies, and do it with far fewer resources than anyone expected. These machines aren’t here yet, but the crypto market reacted anyway, with prices, including BTC/USD, dipping.
Earlier, Google unveiled TurboQuant, a new compression method that could cut the memory requirements of large language models by up to six times. The news sent ripples through the markets, with memory chip makers in Asia and the U.S. seeing their shares fall, along with the Nasdaq index, amid fears that demand could weaken.
Still, these worries might be overblown. We’ve seen similar reactions before. Last year, DeepSeek was hyped as far more efficient than ChatGPT, and Nvidia stock plunged more than 15% in a single day, wiping out over $500 billion in market value. Yet a year later, chip demand is still strong, showing that hype doesn’t always change the bigger picture.
In fact, Nvidia’s CEO, Jensen Huang, now expects the company to generate at least $1 trillion in revenue from its Blackwell and Rubin chips by the end of 2027 — double his previous forecast of $500 billion for 2026. AMD is also seeing stronger-than-expected demand for server CPUs, fueled largely by the boom in so-called “agent” AI applications.
Still, there are real risks. If tensions in the Strait of Hormuz remain high, keeping energy prices elevated and disrupting supplies of critical materials like helium, which accounts for about 30 percent of global supply from Qatar, production costs could rise. If that ripple spreads through the broader economy, it could weigh on demand.
The good news is that the U.S. and Iran appear to be in talks to ease tensions. At the same time, additional military deployments in the region suggest that further escalation and a longer-lasting shock to the semiconductor industry cannot be ruled out.
The bottom line is that risks are rising, but there’s no clear sign the “bubble” is about to burst. The key indicator to watch is hyperscalers’ spending on data centers. If that slows, for example, if AI doesn’t deliver the returns everyone is expecting, it could pull the whole market down more sharply than current geopolitical tensions.