Nvidia’s chips fuel A.I. — Here’s why the U.S. worries about China’s access
Jensen Huang, president and CEO of Nvidia, speaks during the Computex Show in Taipei on May 30, 2017.
SAM YEH | AFP | Getty Images
Nvidia’s powerful semiconductors have taken on particular importance as their capacity to fuel artificial intelligence has become increasingly sought-after.
But their unique capacity is also what’s made China hawks in the U.S. fearful about what it could mean for them to get into the wrong hands, where it could be used to accelerate the spread of non-democratic ideas or develop autonomous weapons.
“If the democratic side is not in the lead on the technology, and authoritarians get ahead, we put the whole of democracy and human rights at risk,” Eileen Donahoe, a former U.S. ambassador to the U.N. Human Rights Council and now executive director of Stanford University’s Global Digital Policy Incubator, told NBC in a recent interview.
With U.S. AI executives warning the government that China is not far behind in its development of the transformative technology, U.S. policymakers believe there’s deep urgency in taking steps to stay ahead.
That’s why the Commerce Department is reportedly considering new limits on the export of such chips to China, The Wall Street Journal reported on Tuesday. Nvidia had already created a version of the A100, its popular AI chip, that it could sell to the Chinese market. The A800 was made to stay within performance parameters previously outlined by the Commerce Department.
But the new limits reportedly being considered by the Biden administration would restrict even those sales without a license.
Such a move would continue the ongoing standoff between the U.S. and Chinese governments on technology sales between the two countries. The Chinese government in May barred “critical information infrastructure” from buying products from U.S. memory chipmaker Micron, saying the company poses a “major security risk.” The U.S. Commerce Department at the time said they “firmly oppose restrictions that have no basis in fact.”
U.S. limitations on the sale of chips with AI capacity to China would make it harder for China to keep up with the pace of development of the sector by U.S. companies like Google and Microsoft-backed OpenAI. While Chinese companies may have some additional advanced chips saved up or resort to slower semiconductors, further limits on high-speed chips could limit their agility in the AI race.
American executives have warned the U.S. government that AI produced without the proper guardrails can be used in nefarious ways. AI models can be designed in ways that perpetuate discrimination in contexts including law enforcement actions, housing and loan approvals and more, for example. But they can also be used to mass produce convincing propaganda or even to develop autonomous weapons.
The Commerce Department did not immediately respond to CNBC’s request for comment on the Journal report and Nvidia declined to comment.
-CNBC’s Kristina Partsinevelos contributed to this report.
WATCH: Can China’s ChatGPT clones give it an edge over the U.S. in an A.I. arms race?
Amazon on Wednesday unveiled a new robotic system that’s capable of performing multiple tasks at once in the company’s warehouses.
The system, called Blue Jay, is made up of a series of robotic arms that are suspended from a conveyor belt-like track. Those arms are tipped with suction-cup devices that allow them to grab items of varying shapes and sizes.
Blue Jay combines “what used to be three separate robotic stations into one streamlined workplace that can pick, sort, and consolidate in a single place,” Amazon said in a blog.
The robotic system’s goal is to assist employees with otherwise strenuous tasks “while creating greater efficiency in less physical space,” the company said.
Amazon is testing Blue Jay at one of its warehouses in South Carolina. So far, the company has observed that the system is able to pick, pack, stow and consolidate “approximately 75% of items we store at our sites.”
Blue Jay joins a growing fleet of robotic machinery being deployed across Amazon’s legions of warehouses. Over the past several years, Amazon has debuted robots capable of handling different tasks, ranging from removing items from shelves to sorting boxes. In May, it debuted “Vulcan,” a robotic system that has a sense of touch.
Amazon’s warehouse automation efforts were largely jumpstarted by its $775 million acquisition of Kiva Systems in 2012.
The announcement comes as Amazon’s warehouse automation has come under growing scrutiny, particularly over how the technology is impacting its sprawling frontline workforce.
The New York Times on Tuesday published an investigation showing that Amazon’s automation team expects that it can avoid hiring more than 160,000 people in the U.S. by 2027, amounting to savings of about 30 cents on every item that Amazon packs and delivers. The report was based on interviews and internal strategy documents, the Times said.
In response to the report, an Amazon spokesperson told CNBC that the documents offer an “incomplete and misleading picture of our plans.”
“In this instance, the materials appear to reflect the perspective of just one team and don’t represent our overall hiring strategy across our various operations business lines — now or moving forward,” the spokesperson said in an email.
As the nation’s second-largest private employer, Amazon’s automation playbook could become a bellwether for the broader job market and other corporations. The company had more than 1.54 million employees globally at the end of the second quarter. That figure excludes delivery drivers, which are contracted through third-party firms.
The company on Wednesday said that employees remain “at the center” of its robotics development. Amazon said its goal is to “reduce physically demanding tasks, simplify decisions and open new career opportunities” for workers.
Amazon has sought to highlight how increasing automation in its facilities will lead to employees adopting “more rewarding” roles within the company. It offers an apprenticeship program in mechatronics and robotics, which involves honing skills around maintaining and monitoring robotic machinery.
WATCH: Meet Vulcan, Amazon’s new stowing robot that can feel what it touches
Applied Digital said on Wednesday that it signed a $5 billion infrastructure lease agreement with a U.S. hyperscaler.
Shares of the data center company dropped more than 7% following the announcement, continuing a recent slumped that’s sent the stock down over 20% in the past week. The stock has still almost quadrupled this year.
The lease announced on Wednesday is for about 15 years and will deliver 200 megawatts of capacity at the company’s Polaris Forge 2 campus in North Dakota. It brings the company’s total leased capacity to 600 megawatts at its two Polaris Forge campuses.
Across the tech industry, the major cloud providers and other internet giants are rapidly investing in artificial intelligence infrastructure and announcing plans for massive new data centers to handle an expected surge in demand. Applied Digital didn’t name its partner for the latest agreement, just disclosing that it’s an “investment grade hyperscaler.”
In an interview with CNBC’s “Squawk on the Street,” CEO Wes Cummins said the five U.S. hyperscalers are Microsoft, Meta, Oracle, Amazon and Google, “so that’s really who we’re targeting.” He said the tenant for the first lease was CoreWeave.
“We started down this path a couple years ago and we stubbed our toe a few times, but I think we’ve really dialed in the process of the ability to build at scale,” Cummins said, adding that the company has a 4 gigawatt “active pipeline.”
In June, Applied Digital announced two long-term lease agreements with CoreWeave for 250 megawatts of capacity. The company said it expects $7 billion in rental revenue over 15 years, and the shares soared 48% on that news.
Applied Digital also secured $5 billion in infrastructure funding from Macquarie Asset Management earlier this month.
“We believe Polaris Forge 2 builds on that momentum, reflecting the strength of our partnerships and the speed at which we’re reshaping the AI infrastructure landscape,” Cummins said in Wednesday’s release.
-
Sports3 years ago‘Storybook stuff’: Inside the night Bryce Harper sent the Phillies to the World Series
-
Sports2 years agoStory injured on diving stop, exits Red Sox game
-
Sports2 years agoGame 1 of WS least-watched in recorded history
-
Sports3 years agoButton battles heat exhaustion in NASCAR debut
-
Sports3 years agoMLB Rank 2023: Ranking baseball’s top 100 players
-
Sports4 years ago
Team Europe easily wins 4th straight Laver Cup
-
Environment2 years agoJapan and South Korea have a lot at stake in a free and open South China Sea
-
Environment1 year agoHere are the best electric bikes you can buy at every price level in October 2024