Two technology booms -- some people might call them frenzies -- are combining to turn a once-obscure type of microprocessor into a must-have but scarce commodity. Artificial-intelligence systems, made by companies ranging in size from Google to the Chinese startup Malong Technologies, rely heavily on a computer chip called a graphics processing unit, or GPU. The chips are also very useful in mining digital currencies like Ethereum, a bitcoin alternative riding the same wave of hype as its more famous cousin. With people and companies involved in the two surging tech niches buying up the same chips, GPUs have been in short supply over the past several months. Prices have increased by as much as 50 percent, according to some resellers and customers. "The chips are simply going out of stock," said Matt Scott, a technologist from the United States who founded Malong after leaving Microsoft's research lab in Beijing in 2014.

Malong, which is based in Shenzhen, China, is building a system that can analyze digital photos and learn to recognize objects. Doing so requires an enormous number of photos, and analyzing all these photos depends on the G.P.U. When the company recently ordered new hardware from a supplier in China, the shipment was delayed by four weeks. And the price of the chips was about 15 percent higher than it had been six months earlier. "We need the latest G.P.U.s to stay competitive," Mr. Scott said.