Latest news with #post-ChatGPT
Bloomberg
a day ago
- Business
- Bloomberg
Prime Day Shoppers Scour Amazon and Rivals for the Best Price
Welcome to Tech In Depth, our daily newsletter about the business of tech from Bloomberg's journalists around the world. Today, Spencer Soper reports on what we can interpret from the early purchases during Amazon's Prime Day sales event. The AI boom goes on: Sales at Taiwan Semiconductor Manufacturing Co. jumped 39% in the June quarter, surprising analysts, and suggesting no slowdown in the post-ChatGPT boom in artificial intelligence spending.
Mint
2 days ago
- Business
- Mint
Nvidia's Huang to Meet Chinese Leaders While AI Curbs Deepen
(Bloomberg) -- Nvidia Corp. co-founder Jensen Huang will meet with senior Chinese officials in Beijing next week, signaling the company's commitment to a vast market Washington is increasingly seeking to isolate. The chief executive officer is seeking discussions with leaders including the commerce minister, a person familiar with the situation said. Huang is planning those meetings while attending the International Supply Chain Expo in Beijing next week, the person said, asking to remain anonymous discussing a plan still in flux. That conference is one of the Chinese government's signature events, and has featured the likes of Apple Inc.'s Tim Cook in the past. Huang, who's been vocal about the need for US companies to access the world's largest semiconductor market, is a frequent visitor to China. He's returning to the country at a sensitive time for the company, which has become ensnared in a broader US-China tech conflict as the foremost producer of chips for AI development. It's unclear what Huang intends to address with Chinese officials. Nvidia representatives declined to comment on his agenda. A commerce ministry spokesperson said the agency had no information to share, when asked about Huang's visit. A representative for the conference organizers declined to comment. The Financial Times reported earlier on Thursday that Huang planned to meet top officials during the expo in Beijing. Nvidia's CEO this year branded Washington's efforts to stall Beijing's semiconductor ambitions a failure, arguing that the US should ease technology export curbs because they hand local rivals like Huawei Technologies Co. an unfair advantage. The company is now barred from selling all but its lower-end, gaming-focused graphics processors in China. Any relaxing of restrictions would benefit Nvidia. It made history this week as the first company to hit $4 trillion of market value, a testament to its central role in providing the hardware for a post-ChatGPT AI infrastructure building boom. Still, Washington remains intent on pursuing a campaign to choke off China's access to cutting-edge technology. The Trump administration has drafted plans to restrict shipments of AI chips to Malaysia and Thailand, part of an effort to crack down on suspected semiconductor smuggling into China. Nvidia said in May — before the latest curbs — it expects to lose out on $8 billion of sales this quarter because of US restrictions generally. It plans to design and sell a new, lower-end AI chip for China this year that won't run afoul of those regulations, the Financial Times reported. More stories like this are available on
Business Times
2 days ago
- Business
- Business Times
TSMC revenue climbs 39% in latest sign of AI spending boom
[TAIPEI] Taiwan Semiconductor Manufacturing Company's (TSMC) revenue rose a better-than-anticipated 39 per cent in the June quarter, buoying expectations for a sustained post-ChatGPT boom in artificial intelligence (AI) spending. The chipmaker for Nvidia and Apple sales climbed to NT$934 billion (S$41 billion) for the three months, based on its reported monthly revenue. That beat the average analyst projection for about NT$928 billion. Investors have piled back into AI-linked companies, shaking off a funk that settled in after China's DeepSeek cast doubt on whether the likes of Meta Platforms and Google needed to spend that much money on data centres. This week, Nvidia became the first company in history to hit a US$4 trillion valuation, underscoring investors' renewed enthusiasm for companies such as TSMC, key to building the infrastructure for AI. TSMC chief executive officer Wei CC reassured shareholders in June that AI chip demand still outstripped supply, and reaffirmed an outlook for 2025 sales to grow in the mid-20 per cent range in US dollar terms. His company has pledged to spend another US$100 billion ramping up manufacturing in Arizona, in addition to an expansion in Japan, Germany and back home. As the world's largest contract chipmaker, TSMC sits at the heart of the global technology supply chain, producing cutting-edge chips for iPhones and Nvidia's AI offerings. While Nvidia is fuelling its growth, TSMC remains reliant on Apple and smartphone makers for most of its business. For 2025, investors remain wary about the impact of tariffs on the global economy and the electronics sector. The Trump administration's trade war is prompting economists to scale back their forecasts for gross domestic product growth worldwide, casting doubt over the outlook for everything from iPhone demand to computing. BLOOMBERG
&w=3840&q=100)
Business Standard
5 days ago
- Business
- Business Standard
Samsung's profit halves, falls 1st time since 2023 on chip business crisis
The disappointing results underscore how South Korea's largest company has ceded leadership in the AI market to SK Hynix Inc. in the post-ChatGPT infrastructure boom Bloomberg By Yoolim Lee Samsung Electronics Co.'s profit fell for the first time since 2023, reflecting the deepening market share losses clouding the memory chipmaker's prospects in the AI era. The company reported a sharper-than-anticipated 56 per cent plummet in operating income for the June quarter, which it blamed on inventory writedowns following US curbs on Chinese-bound AI chips. Its shares swung between gains and losses in Seoul Tuesday after the company announced a 3.9 trillion won stock buyback. The disappointing results underscore how South Korea's largest company has ceded leadership in the AI market to SK Hynix Inc. in the post-ChatGPT infrastructure boom. Its longstanding rival — along with Micron Technology Inc. — now sells more of the cutting-edge high-bandwidth memory chips paired with Nvidia Corp.'s AI accelerators. Compounding the problem, US restrictions on tech exports to China are hindering a turnaround in its loss-making contract chipmaking operation. Still, some investors expect the chips-to-smartphone company to have hit bottom over the summer. Nvidia is shifting toward a new generation of memory chips, offering Samsung an opening. The one-time inventory adjustment suggests it wants to start the second half with a clean slate, said Sanjeev Rana, head of research at CLSA Securities Korea. 'This is likely to be the bottom,' he said. 'Looking beyond the disappointing results in the second quarter, we expect a sequential recovery.' The company reported a preliminary operating profit of 4.6 trillion won ($3.3 billion) in the June quarter — the company's lowest since 2023 and short of analysts' projections. Revenue was flat at 74 trillion won. The company will provide a full financial statement with net income and divisional breakdowns later this month. Profit fell after Samsung's foundry arm, which has relied in part on Chinese demand, booked a one-time inventory cost on unsold AI chips. Usage rates also fell, the company said in an unusual statement issued to explain the worse-than-expected performance. Operating losses in its contract chipmaking business are expected to narrow in the second half of the year on a gradual recovery in demand, Samsung said. The key for Samsung now is next-generation AI memory. Attention has been on Samsung's attempt to secure certification from Nvidia for its most advanced product, 12-layer HBM3E. Its failure to do so is creating an unusually long lead time for SK Hynix in the highly lucrative space, while US competitor Micron has been rapidly advancing, further reducing Samsung's pie. Customer evaluation and shipments of its advanced memory products are proceeding, Samsung said. What Bloomberg Intelligence Says Samsung could lag SK Hynix in operating profit during 2H as it reported 2Q operating profit 26 per cent below consensus. Its high-bandwidth memory (HBM) chips could remain affected by US regulations which restrict exports to China. Also, US tariffs on consumer electronics and smartphones might weigh more on Samsung than on SK Hynix. But Samsung's 3Q profit could recover sequentially as seasonality kicks in. -Masahiro Wakasugi, Takumi Okano, BI analysts SK Hynix has meanwhile aggressively positioned itself as Nvidia's primary HBM4 supplier. It shipped the world's first 12-layer HBM4 samples to customers ahead of schedule, followed by Micron in June, while Samsung has had to revise its 12-layer HBM3E design. Bernstein analysts led by Mark Li estimated in a June 23 note that SK Hynix holds 57 per cent of the HBM market in 2025, followed by Samsung at 27 per cent and Micron at 16 per cent. At its annual shareholder meeting in March, Samsung vowed to strengthen its position in the HBM market this year, responding to concerns over its underperformance in AI. Jun Young-hyun, head of Samsung's chip business, said that Samsung's failure to secure an early lead in the HBM market contributed to it lagging behind rival SK Hynix and pledged not to repeat the mistake with HBM4. The next-generation memory is expected to be used in Nvidia's Rubin GPU architecture. In April, Samsung said it shipped enhanced HBM3E samples to major customers and expected that product line to contribute to revenue in the second quarter. The company had also said it plans to begin mass production of HBM4 chips in the second half of the year.
Yahoo
5 days ago
- Business
- Yahoo
AI is forcing the data industry to consolidate — but that's not the whole story
The data industry is on the verge of a drastic transformation. The market is consolidating. And if the deal flow in the past two months is any indicator — with Databricks buying Neon for $1 billion and Salesforce snapping up cloud management firm Informatica for $8 billion — momentum is building for more. The acquired companies may range in size, age, and focus area within the data stack, but they all have one thing in common. These companies are being bought in hopes the acquired technology will be the missing piece needed to get enterprises to adopt AI. On the surface level, this strategy makes sense. The success of AI companies, and AI applications, is determined by access to quality underlying data. Without it, there simply isn't value — a belief shared by enterprise VCs. In a TechCrunch survey conducted in December 2024, enterprise VCs said data quality was a key factor to make AI startups stand out and succeed. And while some of these companies involved in these deals aren't startups, the sentiment still stands. Gaurav Dhillon, the former co-founder and CEO of Informatica, and current chairman and CEO at data integration company SnapLogic, echoed this in a recent interview with TechCrunch. 'There is a complete reset in how data is managed and flows around the enterprise,' Dhillon said. 'If people want to seize the AI imperative, they have to redo their data platforms in a very big way. And this is where I believe you're seeing all these data acquisitions, because this is the foundation to have a sound AI strategy.' But is this strategy of snapping up companies built before a post-ChatGPT world the way to increase enterprise AI adoption in today's rapidly innovating market? That's unclear. Dhillon has doubts too. 'Nobody was born in AI; that's only three years old,' Dhillon said, referring to the current post-ChatGPT AI market. 'For a larger company, to provide AI innovations to re-imagine the enterprise, the agentic enterprise in particular, it's going to need a lot of retooling to make it happen.' The data industry has grown into a sprawling and fragmented web over the past decade — which makes it ripe for consolidation. All it needed was a catalyst. From 2020 through 2024 alone, more than $300 billion was invested into data startups across more than 24,000 deals, according to PitchBook data. The data industry wasn't immune to the trends seen in other industries like SaaS where the venture swell of the last decade resulted in numerous startups getting funded by venture capitalists that only targeted one specific area or were in some cases built around a single feature. The current industry standard of bundling together a bunch of different data management solutions, each with its own specific focus, doesn't work when you want AI to crawl around your data to find answers or build applications. It makes sense that larger companies are looking to snap up startups that can plug into and fill existing gaps in their data stack. A perfect example of this trend is Fivetran's recent acquisition of Census in May — which yes, was done in the name of AI. Fivetran helps companies move their data from a variety of sources into cloud databases. For the first 13 years of its business, it didn't allow customers to move this data back out of said databases, which is exactly what Census offers. This means prior to this acquisition, Fivetran customers needed to work with a second company to create an end-to-end solution. To be clear, this isn't meant to cast shade on Fivetran. At the time of the deal, George Fraser, the co-founder and CEO of Fivetran, told TechCrunch that while moving data in and out of these warehouses seems like two sides of the same coin, it's not that simple; the company even tried and abandoned an in-house solution to this problem. 'Technically speaking, if you look at the code underneath [these] services, they're actually pretty different,' Fraser said at the time. 'You have to solve a pretty different set of problems in order to do this.' This situation helps illustrate how the data market has transformed in the last decade. For Sanjeev Mohan, a former Gartner analyst who now runs SanjMo, his own data trend advisory firm, these types of scenarios are a big driver of the current wave of consolidation. 'This consolidation is being driven by customers being fed up with a multitude of products that are incompatible,' Mohan said. 'We live in a very interesting world where there are a lot of different data storage solutions, you can do open source, they can go to Kafka, but the one area where we have failed is metadata. Dozens of these products are capturing some metadata but to do their job, it's an overlap.' The broader market plays a role here too, Mohan said. Data startups are struggling to raise capital, Mohan said, and an exit is better than having to wind down or load up on debt. For the acquirers, adding features gives them better pricing leverage and an edge against their peers. 'If Salesforce or Google isn't acquiring these companies, then their competitors likely are,' Derek Hernandez, a senior emerging tech analyst at PitchBook, told TechCrunch. 'The best solutions are being acquired currently. Even if you have an award-winning solution, I don't know that the outlook for staying private ultimately wins over going to a larger [acquirer].' This trend brings big benefits to the startups getting acquired. The venture market is starving for exits and the current quiet period for IPOs doesn't leave them a lot of opportunities. Getting acquired not only provides that exit, but in many cases gives these founding teams room to keep building. Mohan agreed and added that many data startups are feeling the pains of the current market regarding exits and the slow recovery of venture funding. 'At this point in time, acquisition has been a much more favorable exit strategy for them,' Hernandez said. 'So I think, kind of both sides are very incentivized to get to the finish line on these. And I think Informatica is a good example of that, where even with a bit of a haircut from where Salesforce was talking to them last year, it's still, you know, was the best solution, according to their board.' But the doubt still remains if this acquisition strategy will achieve the buyers' goals. As Dhillon pointed out, the database companies being acquired weren't necessarily built to easily work with the rapidly-changing AI market. Plus, if the company with the best data wins the AI world, will it make sense for data and AI companies to be separate entities? 'I think a lot of the value is in merging the major AI players with the data management companies,' Hernandez said. 'I don't know that a standalone data management company is particularly incentivized to remain so and, kind of like, play a third party between enterprises and AI solutions.' 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
