Latest news with #Computex
South China Morning Post
15 hours ago
- Business
- South China Morning Post
Trade war elephant in the showroom at Computex Taipei, Asia's biggest tech expo
There was an elephant in the showroom at this year's Computex Taipei tech expo. Advertisement Loud speakers boomed the guttural sounds of action games loud enough to be heard across dozens of booths, where blazing white lights glinted off the thin black edges of new microprocessors on display. Some of the show's 86,521 prospective buyers tried out AI-enabled laptops with 3D displays – visible without special glasses – screening a country lane in Japan one moment, a complicated colon surgery the next. And graphics card developer Nvidia's CEO, Jensen Huang, talked big about building a 'supercomputer' for start-ups in Taiwan , where the show took place. But the scene differed little from Computex 2024, held 18 months after OpenAI's game-changing release of ChatGPT. By June last year, that window had given developers enough time to release foundational hardware, such as AI computing chips and accelerators for cloud servers – a boost to Taiwan's economic growth, including an unexpectedly fast 5.37 per cent year-on-year uptick in the first quarter of 2025. Advertisement Instead, this year's show rolled out incremental improvements, such as mouse clicks that activate AI software and microphones that allow users to issue commands. Huang did not hit the exhibition floor this year to sign autographs, unlike in 2024 when wave after wave of show-goers queued at the larger booths to sample AI-powered laptops. But in the grand scheme of things, that was just a bug. Now, the elephant.
Business Insider
a day ago
- Business
- Business Insider
Nvidia's most important number doesn't have a dollar sign in front of it: Token growth
Nvidia 's revenue climbed to $44.1 billion last quarter, yet one of the chip giant's most important metrics doesn't have a dollar sign in front of it. "OpenAI, Microsoft, and Google are seeing a step function leap in token generation," said Nvidia CFO Colette Kress on the company's Wednesday earnings call. "Microsoft processed over 100 trillion tokens in Q1, a fivefold increase on a year-over-year basis," she continued. Throughout May, top tech executives — many of whom are also some of Nvidia's largest customers — have been boasting about their token growth. Though somewhat hard to track from outside the AI cloud or foundation model companies themselves, tokens are the base unit to measure AI inputs and outputs. They include pixels, word segments, or audio. But no matter the content, all AI breaks down into tokens. As AI tools mature, the number of tokens generated for AI outputs, or inference, is growing faster than many expected. When Google CEO Sundar Pichai said monthly tokens produced across Google's products had increased by a factor of 50 in the last year, the Google I/O audience gasped. "Explosive token growth is what really matters, in the longer term," wrote Morgan Stanley analysts ahead of Nvidia's latest earnings call, after which Nvidia's share price climbed to $134. Why tokens matter Forrester analyst Alvin Nguyen explained that it's not a perfect metric, since tokens can vary in size based on the form of content they represent. "It isn't a clear way to make apples-to-apples comparisons, but it is the closest to a standard that we have without needing more data and analytics," Nguyen said. Nvidia CEO Jensen Huang sees the rise of the "token" in the conversations of top tech executives as a sign that AI tools are providing value. "Where companies are starting to talk about how many tokens they produced last quarter and how many tokens they produced last month. Very soon we'll be talking about how many tokens we produce every hour, just as every single factory does," Huang said at Computex in Taiwan last week. With its dominant market share, Nvidia directly benefits from almost all token growth. Some analysts say the demand is growing faster than the current data center stock can handle. "Every hyperscaler has reported unanticipated strong token growth," the Morgan Stanley analysts wrote, adding that plentiful anecdotal evidence of more inference demand than existing infrastructure can support strengthened their conviction. The problem with tokens is that there's no easy way to gauge their growth unless companies release numbers. But it's safe to say that when executives share any indication of demand for tokens, investors will be paying close attention for the foreseeable future.
CNA
2 days ago
- Business
- CNA
Commentary: Consumer AI gadgets will come with a whimper, not a bang
TOKYO: Where are all the artificial intelligence consumer gadgets? Even a year ago, it seemed tech companies were working to incorporate the technology into every physical device, from coffee makers to vacuums, making 'AI-powered' hardware seem like it would soon be as ubiquitous as 'battery-powered' electronics. Typically, tech conferences offer a glimmer of these futuristic toys. Not all of them end up hitting the market, but it's where we can dream a little about new pocket devices or household robots taking on a greater role in our lives. So it was a little disappointing last week at Asia's biggest artificial intelligence conference, Taiwan's Computex, to find hardly any mentions of consumer-facing tech. Most keynotes focused on enterprise applications of AI, such as agents or automated manufacturing. Walking around the exhibitors' hall, the only thing that caught my eye were wireless computer mice shaped like cats. LACK OF NEW AI GADGETS A few things seem to have changed. For starters, there's the reality that hardware engineering presents an entirely different set of physical challenges compared to tinkering with AI software. And a global trade war also makes it a risky time to launch a new gadget when it's unclear if consumers are interested. Companies also may be starting to pick up on the fact that while Wall Street is awash with global hype on the AI boom, it isn't exactly a selling point on Main Street. If anything, some of the executives speaking at the conference threw cold water on the next generations of these AI-first consumer products. Asustek Computer co-Chief Executive Officer Samson Hu told Bloomberg News that it will take another year or more for AI-powered PCs to go mainstream because the technology isn't quite there yet and macroeconomic uncertainty is impacting people's spending. There have been few compelling use cases for AI PCs so far, despite the mountain of promotion. Meanwhile, the graveyard of AI hardware that was supposed to transform our lives is already growing. The Humane AI Pin wearable device – launched last year to much hype about how it was going to replace the smartphone – ended up receiving brutal reviews while being a fire hazard. The startup, run by two former Apple employees, stopped selling the AI Pin earlier this year and was sold for parts. The Rabbit R1 assistant is another cautionary tale of the false promises of these gadgets. But that doesn't mean the future of AI consumer products isn't coming. OpenAI made the major announcement last week that it is working with legendary iPhone designer Jony Ive to launch something that takes AI into the physical realm for consumers. But even the might of OpenAI's technology and Ive's design prowess make whatever it is a far from certain bet. TACKLING TASKS ONE-BY-ONE There were perhaps some lessons for the future of such devices from the gathering in Taiwan. During his keynote speech, Qualcomm Chief Executive Officer Cristiano Amon said that AI computers are at the phase where they will require the work of outside developers to make them appeal to consumers. The iPhone, for example, didn't take off immediately after it was launched. But it became essential to so many people because of the myriad apps developers built on top of it that we now use to hail taxis, order food or move around new cities. 'Really, the developer ecosystem is going to make this shift to AI PCs,' Amon said. He's right, and the same is true beyond just AI computers. For any revolutionary AI hardware device, and especially a smartphone killer, the more that global developers lead the charge to meet peoples' needs and solve small, everyday problems, the more likely they are to succeed. In this economy, maybe that doesn't mean repackaging the same old gadgets with shiny new AI labels. It means iterating and perfecting real use cases that incorporate the buzzy technology into devices and make our lives easier. And this will inevitably be a collective effort. AI is already transforming our world in small ways. I find asking ChatGPT to quickly translate phrases for me while on the go a lifesaver when navigating a new country. But I hardly want to shell out money to carry around a new device simply to access ChatGPT. The more the tech industry tries to slap AI onto everything and market it as a panacea for all our problems, the more I get a snake-oil salesman ick.
Yahoo
2 days ago
- Business
- Yahoo
Nvidia Aims To Bypass Big Tech Dependency By Going Straight To International Governments, As It Did Recently With Saudi Arabia
Nvidia (NASDAQ:NVDA) is looking beyond Big Tech as it strategizes global expansion by doing deals with foreign governments, the Financial Times reports. As the innovative chip company behind artificial intelligence, demand for Nvidia's technology crosses all boundaries — from neighboring Big Tech companies in Silicon Valley to foreign governments. Following Nvidia's recently announced multibillion-dollar chip agreement with as a part of a deal with the U.S. government to build a large AI infrastructure, the FT reports that such 'sovereign AI' deals are a part of Nvidia's strategy to be less reliant on 'hyperscalers' — large cloud computing companies accounting for over half its data center revenue. Don't Miss: Hasbro, MGM, and Skechers trust this AI marketing firm — Maker of the $60,000 foldable home has 3 factory buildings, 600+ houses built, and big plans to solve housing — To do this, Nvidia is allowing access to its technology from up-and-coming rivals to enable them to foster relationships with cloud service providers like Amazon (NASDAQ:AMZN) and Google, FT reports. The outlet adds that business relationships with Cisco (NASDAQ:CSCO), Dell (NYSE:DELL), and Hewlett-Packard Enterprise Company (NASDAQ:HPE) will help them enable enterprise customers to create their own IT infrastructure instead of outsourcing to the cloud. 'I'm more certain [about the business opportunity beyond the big cloud providers] today than I was a year ago,' Nvidia CEO Jensen Huang told the FT in March. Trending: 'Scrolling To UBI' — Deloitte's #1 fastest-growing software company allows users to earn money on their phones. As part of its expansion, Nvidia sees robots as pivotal to introducing AI to practical real-world uses in robotics, drug development and autonomous navigation. The company presented employee-authored literature at a Singapore conference covering these sectors, highlighting the partnerships with Google, GE HealthCare (NASDAQ:GEHC), and General Motors (NYSE:GM), FT said. Experts question whether the scope of the deals Nvidia just made in Saudi Arabia is feasible on an ongoing basis. 'Is every country going to announce a $10 billion or $50 billion data center like the Saudis? Of course not,' Seaport Research analyst Jay Goldberg told Reuters. 'They're sort of running out of obvious deals.' Huang, however, remains bullish: 'AI infrastructure is being built out (everywhere) – that's one of the reasons I'm traveling around the world... AI infrastructure is going to be a part of society,' he told explained at the recent Computex conference in Taipei that new technology, rather than mega infrastructure deals with sovereign governments, was the key to the company's expansion, Reuters reported. Specifically, Huang spoke about NVLink Fusion, which allows companies to access Nvidia's AI infrastructure using custom chips, creating a scalable architecture. Nvidia's NVLink Fusion platform will encourage companies to build compliant hardware and thus drive demand for Nvidia AI network and data centre parts, Reuters says. It means that Nvidia can be a crucial part of a country or company's growth without having to carry all the weight of expansion. Read Next:Inspired by Uber and Airbnb – Deloitte's fastest-growing software company is transforming 7 billion smartphones into income-generating assets – Image: Shutterstock UNLOCKED: 5 NEW TRADES EVERY WEEK. Click now to get top trade ideas daily, plus unlimited access to cutting-edge tools and strategies to gain an edge in the markets. Get the latest stock analysis from Benzinga? APPLE (AAPL): Free Stock Analysis Report TESLA (TSLA): Free Stock Analysis Report This article Nvidia Aims To Bypass Big Tech Dependency By Going Straight To International Governments, As It Did Recently With Saudi Arabia originally appeared on © 2025 Benzinga does not provide investment advice. All rights reserved. Sign in to access your portfolio
CNBC
2 days ago
- Business
- CNBC
Op-ed: Don't buy Nvidia CEO Jensen Huang's China 'failure' story
When Nvidia CEO Jensen Huang appeared at Computex in Taipei last week, his presence dominated headlines. But it wasn't his keynote that drew political attention — it was what he said after the speech. In interviews on the sidelines of the event, Huang sharply criticized U.S. export controls on AI chips, calling them a "failure." He blamed both the Trump and Biden administrations for triggering a collapse in Nvidia's China business, claiming that the company's market share in the country dropped from 95% to 50% over four years, and disclosed a multi-billion dollar write-down on unsold H20 chips blocked from sale. Huang further argued that these restrictions have only pushed Chinese firms to innovate faster, intensifying competition. The Nvidia CEO's complaints about U.S. policy continued after Wednesday's earnings report, with Huang saying China is "effectively closed." "Export restrictions spurred China's innovation," Huang said during the earnings call. "The U.S. has based its policy on the assumption that China cannot make AI chips. Assumption was always questionable. Now it's clearly wrong." "The question is not whether China will have AI. It already does," he added. On Wednesday, Chinese generative AI company DeepSeek, whose debut was the watershed moment in China's arrival as a competitor to the best the U.S. has to offer in the AI race, released a new model. While Huang's frustrations may resonate with some in the business world, they badly miss the strategic mark. The semiconductor policies Huang criticizes were never designed to protect Nvidia's commercial interests in China. They were, and remain, about U.S. national security. The Biden administration's controls, building on Trump-era efforts, reflect a sober reality: the U.S. cannot afford to fuel the military rise of its chief strategic rival. Advanced chips power the modernization of the People's Liberation Army, including command, control, communications, computers, intelligence, surveillance, and reconnaissance (C4ISR) systems. These restrictions are not about quarterly earnings — they are about avoiding scenarios where American troops face U.S.-designed technology on the battlefield. Critics like Huang claim these controls have backfired, accelerating China's tech independence. But that effort didn't begin with the 2022 restrictions. It was formalized years earlier through the "Made in China 2025" strategy, approved by the State Council in 2015. With over $150 billion in state support, China's chip sector has long aimed to reduce dependence on foreign suppliers. The controls didn't start this race; they were a response to it. In an interview with CNBC's Jim Cramer on "Mad Money" on Wednesday, Huang said 50% of the world's AI developers are in China and "we want the world to build on America's IT stack. ... that's the most important strategic reason to be in China." But as China continues to race against the U.S. in AI, it should not be done by exploiting the U.S. or with U.S. companies willingly helping it. Furthermore, China is making its own decisions regardless of U.S. export controls to favor its own companies, a goal it had well before the Trump or Biden export controls. This may be inconvenient history, but it is history, nonetheless. China's innovation model is evolving. While foreign technology transfer and joint ventures played early roles, Chinese firms now combine global know-how with massive capital, targeted industrial policy, and relentless domestic competition. China ranks second in global R&D spending and leads in green tech, EVs, and high-speed rail. It extracts what it can from foreign firms, and then moves on. The U.S. must safeguard leading-edge technologies and delay diffusion for as long as possible. This is not a permanent advantage, but it is a necessary one. Look at sectors where the U.S. failed to act — solar panels, EV batteries, telecom. In each case, China leapfrogged the West and locked in global dominance. But contrast that with aviation, where the U.S. has maintained strict export controls and upheld a high regulatory bar. China's COMAC, although making noteworthy gains, remains far behind Boeing and Airbus in capability, global market access, safety certifications, and production reliability. Where the U.S. held the line, it held its lead. Nvidia's arc in China follows a familiar pattern. Foreign firms are welcomed, dominate temporarily, and are ultimately displaced. We all know the story by now — Apple, Tesla, Starbucks — all have faced this trajectory. Huang's claim that U.S. policy ended Nvidia's privileged market position ignores this well-established history. The displacement for Nvidia (like for so many others) was coming either way. And despite his warnings, Nvidia has not suffered a collapse. In 2022, Huang warned that export controls could devastate his company. Since then, Nvidia's stock has increased more than tenfold. In fiscal year 2024, its data center revenue jumped 217% year-over-year, driven by soaring demand across the U.S., Europe, and Gulf States. The company's success has continued — despite, or perhaps partially because of, the strategic shift in its market exposure. Its latest results showed overall revenue grew by 69% during the quarter, and in its key data center division — which includes AI chips and related parts — growth was 73% on an annual basis to $39.1 billion. Huang said in an earnings statement that "global demand" continues to be "incredibly strong" for its AI equipment. The real risk is not that these policies have failed — they haven't, when judged against their national security goals. The risk is that their momentum could be disrupted by policy discontinuity, political lobbying, or shifting priorities in Washington. We've already seen the Trump administration roll back one key element: the Biden-era AI diffusion rule. That rule, with a tiered structure aimed at controlling the spread of high-performance chips and AI model weights based on risk level, would have required global licensing and compliance from chipmakers like Nvidia. Its rescission — prompted by commercial and legal concerns — has raised alarms among national security officials. Other export controls, however, remain firmly in place. Nvidia should be using its success around the world to work hard to compete against China in other markets and stop focusing on its diminishing market share in China, especially after Trump gave the company the gift of repelling the AI diffusion rule. Huang may have intended his remarks to influence the policy debate. Instead, they triggered political reaffirmation. In fact, his comments may go down as his "Jack Ma moment" — a high-profile critique swiftly followed by a government response that reasserts authority. Just as Ma's critique of Chinese regulators at the Bund Summit preceded the abrupt suspension of Ant Financial's IPO, Huang's criticism on the margins of Computex was quickly answered. Trump advisor Sriram Krishnan said in an interview with Bloomberg, "the guardrails are staying," emphasizing that AI chips "can be used in ways that fundamentally challenge U.S. national security." On Wednesday, before Nvidia had even reported, shares of chip design companies Cadence and Synopsys sank after the Financial Times reported that the White House told them to stop selling to clients in China. U.S. semiconductor policy isn't about one firm's earnings or market access — it's about protecting America's strategic edge in a high-stakes geopolitical contest. The challenge ahead is not to protect any one company's dominance in China. It's to defend the foundational technologies that define the future of power. Sometimes that means stepping back from markets that were never going to remain open anyway. This is the hard truth of policymaking. Not every decision, few in fact, is about corporate profits. Most are about the safety, security, and long-term stability and strength of the country, and the people who protect it. —
