Nvidia Corp chief executive officer Jensen Huang (黃仁勳) on Tuesday used his annual GPU Technology Conference (GTC) to address concerns that the cost of artificial intelligence (AI) computing is spiraling out of control.
Huang at the GTC unveiled more powerful chips and related technology that he said would provide a clearer payoff to customers. The lineup includes a successor to Nvidia’s flagship AI processor called the Blackwell Ultra, as well as additional generations stretching into 2027.
The new Blackwell Ultra processor lineup would arrive in the second half of this year, and would be followed by a more dramatic upgrade called Vera Rubin in the latter half of next year, and Rubin Ultra in 2027, he said.
Photo: Bloomberg
Vera Rubin’s namesake was an American astronomer credited with helping discover the existence of dark matter.
Huang also unveiled Dynamo-branded software that would fine-tune existing and future equipment, making it more efficient and profitable.
“It is essentially the operating system of an AI factory,” Huang said during a roughly two-hour presentation at the GTC in San Jose, California, which touched on everything from robot technology to personal supercomputers.
The conference, once a little-known gathering of developers, has become a closely watched event since Nvidia assumed a central role in AI — with the tech world and Wall Street taking its cues from the presentation.
Huang introduced a variety of hardware, software and services during his speech, though there were no bombshell revelations for investors. Nvidia shares closed down more than 3 percent on Tuesday.
It is a pivotal moment for Nvidia. After two years of stratospheric growth for its revenue and market value, investors this year have begun to question whether the frenzy is sustainable. These concerns were brought into focus earlier this year when Chinese start-up DeepSeek (深度求索) said it had developed a competitive AI model using a fraction of the resources.
The weeklong GTC is a chance to convince the tech industry that Nvidia chips are still must-haves for AI — a field that Huang expects to spread to more of the economy in what he has called a new industrial revolution.
Dubbed the “Super Bowl of AI,” Huang offered a road map for future chips and unveiled a breakthrough system that relies on a combination of silicon and photonics — light waves.
Nvidia also announced plans for a quantum computing research lab in Boston, aiming to capitalize on another emerging technology.
Other announcements included: A partnership with General Motors Co that would add AI to next-generation cars, factories and robots; a wireless project involving companies such as T-Mobile US Inc and Cisco Systems Inc for new 6G networks; and new Nvidia-based personal supercomputer systems from Dell Technologies Inc, HP Inc and other manufacturers.
Meanwhile, Huang said the top four public cloud vendors — Amazon.com Inc, Microsoft Corp, Alphabet Inc’s Google and Oracle Corp — bought 1.3 million of Nvidia’s older-generation Hopper AI chips last year.
So far this year, the same group has bought 3.6 million Blackwell AI chips, he said, adding that he expects Nvidia’s data center infrastructure revenue to hit US$1 trillion by 2028.
At the end of his presentation, Huang introduced an open-source physics engine for robotics simulation called Newton, which is being developed with Google’s DeepMind and Walt Disney Co.
A small, boxy robot named Blue joined him on stage, popping up from a hatch in the floor. It beeped at Huang and followed his commands, standing beside him as he wrapped up his thoughts.
“The age of generalist robotics is here,” Huang said.
Additional reporting by AP
To many, Tatu City on the outskirts of Nairobi looks like a success. The first city entirely built by a private company to be operational in east Africa, with about 25,000 people living and working there, it accounts for about two-thirds of all foreign investment in Kenya. Its low-tax status has attracted more than 100 businesses including Heineken, coffee brand Dormans, and the biggest call-center and cold-chain transport firms in the region. However, to some local politicians, Tatu City has looked more like a target for extortion. A parade of governors have demanded land worth millions of dollars in exchange
An Indonesian animated movie is smashing regional box office records and could be set for wider success as it prepares to open beyond the Southeast Asian archipelago’s silver screens. Jumbo — a film based on the adventures of main character, Don, a large orphaned Indonesian boy facing bullying at school — last month became the highest-grossing Southeast Asian animated film, raking in more than US$8 million. Released at the end of March to coincide with the Eid holidays after the Islamic fasting month of Ramadan, the movie has hit 8 million ticket sales, the third-highest in Indonesian cinema history, Film
Taiwan Semiconductor Manufacturing Co’s (TSMC, 台積電) revenue jumped 48 percent last month, underscoring how electronics firms scrambled to acquire essential components before global tariffs took effect. The main chipmaker for Apple Inc and Nvidia Corp reported monthly sales of NT$349.6 billion (US$11.6 billion). That compares with the average analysts’ estimate for a 38 percent rise in second-quarter revenue. US President Donald Trump’s trade war is prompting economists to retool GDP forecasts worldwide, casting doubt over the outlook for everything from iPhone demand to computing and datacenter construction. However, TSMC — a barometer for global tech spending given its central role in the
Alchip Technologies Ltd (世芯), an application-specific integrated circuit (ASIC) designer specializing in server chips, expects revenue to decline this year due to sagging demand for 5-nanometer artificial intelligence (AI) chips from a North America-based major customer, a company executive said yesterday. That would be the first contraction in revenue for Alchip as it has been enjoying strong revenue growth over the past few years, benefiting from cloud-service providers’ moves to reduce dependence on Nvidia Corp’s expensive AI chips by building their own AI accelerator by outsourcing chip design. The 5-nanometer chip was supposed to be a new growth engine as the lifecycle
RSS