hyperscale cloud providers

Chipmaker Marvell says custom silicon has become a necessity for cloud service providers as they build out AI infrastructure. The company argues hyperscalers increasingly want tailored chips rather than off-the-shelf processors to optimise performance and cost, a shift that positions Marvell to win more design work from the largest cloud platforms.

Economist Jahangir Aziz says major cloud providers, or hyperscalers, are consuming most of the available semiconductors, leaving security hardware manufacturers struggling with limited supply and rising costs. Companies building firewalls, intrusion detection systems and secure routers face growing difficulties sourcing chips as cloud firms absorb the bulk of global semiconductor output, raising concerns about the availability and price of critical security infrastructure.

Marvell Technology held an investor day where it presented its strategy for custom AI chips to Wall Street, arguing that its bespoke silicon business can capture a growing share of demand as cloud providers design their own accelerators. The event was aimed at convincing investors the company can compete for lucrative custom AI semiconductor contracts alongside larger rivals.

Amazon Web Services' chief executive has publicly flagged what he describes as a surprising risk facing AI-linked stocks and the broader U.S. economy. The comments come as markets weigh heavy corporate spending on AI infrastructure against questions about returns, and investors watch whether hyperscale cloud providers' outlooks could signal a cooling in the AI trade.