Latest news with #MaxA.Cherney
Yahoo
15-07-2025
- Business
- Yahoo
Broadcom launches new Tomahawk Ultra networking chip in AI battle against Nvidia
By Max A. Cherney SAN FRANCISCO (Reuters) -Broadcom's (AVGO) chip unit unveiled on Tuesday a new networking processor that aims to speed artificial intelligence data crunching, which requires stringing together hundreds of chips that work together. The new chip is the latest piece of hardware that Broadcom has brought to bear against rival AI giant Nvidia (NVDA). Broadcom helps Alphabet's Google (GOOG) produce its AI chips, which are perceived by developers and industry experts as one of the few viable alternatives to Nvidia's powerful graphics processors (GPUs). Dubbed the Tomahawk Ultra, Broadcom's chip acts as a traffic controller for data whizzing between dozens or hundreds of chips that sit relatively closely together inside a data center, such as inside a single server rack. The chip aims to compete with Nvidia's NVLink Switch chip which has a similar purpose, but the Tomahawk Ultra can tie together four times the number of chips, Ram Velaga, a Broadcom senior vice president, told Reuters in an interview. And instead of a proprietary protocol to move the data, it uses a boosted-for-speed version of ethernet. Both companies' chips help data center builders and others tie as many chips as possible together within a few feet of each other, a technique the industry calls "scale-up" computing. By ensuring close-by chips can communicate with each other quickly, software developers can summon the computing horsepower necessary for AI. Taiwan Semiconductor Manufacturing will manufacture the Ultra line of processors with its five nano-meter process, Velaga said. The processor is now shipping. It took Broadcom's teams of engineers roughly three years to develop the design, which was originally built for a segment of the market known as high-performance computing. But as generative AI boomed, Broadcom adapted the chip for use by AI companies because it is suited to scaling up. Error while retrieving data Sign in to access your portfolio Error while retrieving data Error while retrieving data Error while retrieving data Error while retrieving data
Yahoo
11-07-2025
- Business
- Yahoo
US senators warn Nvidia CEO about upcoming China trip
By Max A. Cherney SAN FRANCISCO (Reuters) -A bipartisan pair of U.S. senators sent a letter to Nvidia CEO Jensen Huang on Friday about an upcoming trip to China, warning the CEO to refrain from meeting with companies that are suspected of undermining U.S. chip export controls. The letter from Republican Senator Jim Banks and Democratic Senator Elizabeth Warren asked Huang to also abstain from meeting with representatives of companies that are working with the People's Republic of China's military or intelligence bodies and are named on the U.S. restricted export list. "We are worried that your trip to the PRC could legitimize companies that cooperate closely with the Chinese military or involve discussing exploitable gaps in U.S. export controls," the senators wrote. Huang planned to visit China on Friday. An Nvidia spokesperson said, "American wins" when its technology sets "the global standard," and that China has one of the largest bodies of software developers in the world. AI software "should run best on the U.S. technology stack, encouraging nations worldwide to choose America," the spokesperson said. In May at the Computex trade show in Taipei, Huang praised President Donald Trump's decision to scrap some artificial intelligence chip export controls and described the prior diffusion rules as a failure. U.S. restrictions in April on AI chips Nvidia modified to comply with export controls to China would reduce Nvidia's revenue by $15 billion, the CEO said. The hardware necessary to power advanced AI is now subject to a bipartisan consensus related to the free export of such hardware, the senators wrote. Advanced AI hardware could "accelerate the PRC's effort to modernize its military," the letter reads. U.S. lawmakers have grown increasingly concerned about efforts to circumvent export controls to China and proposed a law that would force AI chip companies to verify the location of their products. Last month, Reuters reported that a senior U.S. official said the AI firm DeepSeek is aiding China's military and intelligence operations, and sought to use shell companies to circumvent U.S. AI chip export controls to China. Nvidia is planning to launch a cheaper version of its flagship Blackwell AI chips for China, Reuters reported in May.
Yahoo
11-07-2025
- Business
- Yahoo
US senators warn Nvidia CEO about upcoming China trip
By Max A. Cherney SAN FRANCISCO (Reuters) -A bipartisan pair of U.S. senators sent a letter to Nvidia CEO Jensen Huang on Friday about an upcoming trip to China, warning the CEO to refrain from meeting with companies that are suspected of undermining U.S. chip export controls. The letter from Republican Senator Jim Banks and Democratic Senator Elizabeth Warren asked Huang to also abstain from meeting with representatives of companies that are working with the People's Republic of China's military or intelligence bodies and are named on the U.S. restricted export list. "We are worried that your trip to the PRC could legitimize companies that cooperate closely with the Chinese military or involve discussing exploitable gaps in U.S. export controls," the senators wrote. Huang planned to visit China on Friday. An Nvidia spokesperson said, "American wins" when its technology sets "the global standard," and that China has one of the largest bodies of software developers in the world. AI software "should run best on the U.S. technology stack, encouraging nations worldwide to choose America," the spokesperson said. In May at the Computex trade show in Taipei, Huang praised President Donald Trump's decision to scrap some artificial intelligence chip export controls and described the prior diffusion rules as a failure. U.S. restrictions in April on AI chips Nvidia modified to comply with export controls to China would reduce Nvidia's revenue by $15 billion, the CEO said. The hardware necessary to power advanced AI is now subject to a bipartisan consensus related to the free export of such hardware, the senators wrote. Advanced AI hardware could "accelerate the PRC's effort to modernize its military," the letter reads. U.S. lawmakers have grown increasingly concerned about efforts to circumvent export controls to China and proposed a law that would force AI chip companies to verify the location of their products. Last month, Reuters reported that a senior U.S. official said the AI firm DeepSeek is aiding China's military and intelligence operations, and sought to use shell companies to circumvent U.S. AI chip export controls to China. Nvidia is planning to launch a cheaper version of its flagship Blackwell AI chips for China, Reuters reported in May. Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data
Time of India
09-07-2025
- Business
- Time of India
Arm estimates a 14-fold increase in data center customers since 2021, company says
By Max A. Cherney SAN FRANCISCO: The number of customers that use Arm-based chips in data centers has grown to 70,000, a 14-fold increase since 2021, the company told Reuters. Under the guidance of Chief Executive Rene Haas , chip technology maker Arm has been working to expand its business into the PC market and has made substantial gains selling its architecture for data center chips as well. Like other chip companies, Arm has benefitted from the frenzy around generative artificial intelligence computing and said a significant portion of its data center growth is due to AI. The company said it had seen a 12-fold spike in startups that are using Arm chips from 2021. Arm's strong outlook arrives as the chip industry faces near term challenges. While chips related to building AI data centers have boomed, other large swaths of the semiconductor market such as PC and mobile sales have remained slow. The company declined to provide annual financial guidance due to trade uncertainty when it released results in May. Chips based on the Arm architecture are known for delivering high performance with low energy consumption, which is one of the reasons it powers just about every mobile phone on the planet. Such performance has been adapted by chip designs for data center processors, which typically consume substantially more energy. The data center market earlier proved difficult for Arm to break into, but the company has more recently been aided by cloud computing giants such as Amazon, Alphabet's Google and Microsoft developing home-grown Arm chips for use in their sprawling infrastructure. Customers often rent Arm-based chips through a cloud computing company such as Amazon's AWS. Amazon has rolled out several generations of its data center processor (CPU) since 2018, including artificial intelligence versions, and added millions Arm-based chips to its cloud computing platform. But SoftBank-owned Arm has made big gains in other areas as well, as it seeks to erode the computing dominance of designs from Advanced Micro Devices and Intel, based on the x86 architecture. The group of developers worldwide working to make apps that run its tech is crucial to a chip company's success. According to the company, Arm has roughly doubled the number of applications since 2021 running on Arm-based machines to 9 million. The developer base working with the company's computing architecture has increased by 1.5 times to 22 million since 2021.
Time of India
02-07-2025
- Business
- Time of India
Exclusive-Intel's new CEO explores big shift in chip manufacturing business
By Max A. Cherney, Jeffrey Dastin and Stephen Nellis SAN FRANCISCO: Intel 's new chief executive is exploring a big change to its contract manufacturing business to win major customers, two people familiar with the matter told Reuters, in a potentially expensive shift from his predecessor's plans. If implemented, the new strategy for what Intel calls its "foundry" business would entail no longer marketing certain chipmaking technology, which the company had long developed, to external customers, the people said. Since taking the company's helm in March, CEO Lip-Bu Tan has moved fast to cut costs and find a new path to revive the ailing U.S. chipmaker. By June, he started voicing that a manufacturing process that prior CEO Pat Gelsinger bet heavily on, known as 18A, was losing its appeal to new customers, said the sources, who spoke on condition of anonymity. To put aside external sales of 18A and its variant 18A-P, manufacturing processes that have cost Intel billions of dollars to develop, the company would have to take a write-off, one of the people familiar with the matter said. Industry analysts contacted by Reuters said such a charge could amount to a loss of hundreds of millions, if not billions, of dollars. Intel declined to comment on such "hypothetical scenarios or market speculation." It said the lead customer for 18A has long been Intel itself, and it aims to ramp production of its "Panther Lake" laptop chips later in 2025, which it called the most advanced processors ever designed and manufactured in the United States. Persuading outside clients to use Intel's factories remains key to its future. As its 18A fabrication process faced delays, rival TSMC's N2 technology has been on track for production. Tan's preliminary answer to this challenge: focus more resources on 14A, a next-generation chipmaking process where Intel expects to have advantages over Taiwan's TSMC, the two sources said. The move is part of a play for big customers like Apple and Nvidia, which currently pay TSMC to manufacture their chips. Tan has tasked the company with teeing up options for discussion with Intel's board when it meets as early as this month, including whether to stop marketing 18A to new clients, one of the two sources said. The board might not reach a decision on 18A until a subsequent autumn meeting in light of the matter's complexity and the enormous money at stake, the person said. Intel declined to comment on what it called rumor. In a statement, it said: "Lip-Bu and the executive team are committed to strengthening our roadmap, building trust with our customers, and improving our financial position for the future. We have identified clear areas of focus and will take actions needed to turn the business around." Last year was Intel's first unprofitable year since 1986. It posted a net loss attributable to the company of $18.8 billion for 2024. The Intel chief executive's deliberations show the enormous risks - and costs - under consideration to move the storied U.S. chipmaker back onto solid footing. Like Gelsinger, Tan inherited a company that had lost its manufacturing edge and fell behind on crucial technology waves of the past two decades: mobile computing and artificial intelligence. The company is targeting high-volume production later this year for 18A with its internal chips, which are widely expected to arrive ahead of external customer orders. Meanwhile, delivering 14A in time to win major contracts is by no means certain, and Intel could choose to stick with its existing plans for 18A, one of the sources said. Intel is tailoring 14A to key clients' needs to make it successful, the company said. AMAZON AND MICROSOFT ON 18A Tan's review of whether to focus clients on 14A involves the contract chipmaking portion of Intel, or foundry, which makes chips for external customers. Regardless of a board decision, Intel will make chips via 18A in cases where its plans are already in motion, the people familiar with the matter said. This includes using 18A for Intel's in-house chips that it already designed for that manufacturing process, the people said. Intel also will produce a relatively small volume of chips that it has guaranteed for and Microsoft via 18A, with deadlines that make it unrealistic to wait for the development of 14A. Amazon and Microsoft did not immediately comment on the matter. Intel said it will deliver on its customer commitments. Tan's overall strategy for Intel remains nascent. So far, he has updated his leadership team, bringing in new engineering talent, and he has worked to shrink what he considered bloated and slow-moving middle management. Shifting away from selling 18A to foundry customers would represent one of his biggest moves yet. The 18A manufacturing process includes a novel method of delivering energy to chips and a new type of transistor. Together, these enhancements were meant to let Intel match or exceed TSMC's capabilities, Intel executives have previously said. However, according to some industry analysts, the 18A process is roughly equivalent to TSMC's so-called N3 manufacturing technology, which went into high-volume production in late 2022. If Intel follows Tan's lead, the company would focus its foundry employees, design partners and new customers on 14A, where it hopes for a better chance to compete against TSMC. Tan has drawn on extensive contacts and customer relationships built over decades in the chip industry to arrive at his view on 18A, the two sources said.
