विज्ञान

By Blake Brittain (Reuters) -The maker of celebrity video platform Cameo sued OpenAI in a California federal court on Tuesday, arguing that the new "Cameo" feature of OpenAI's Sora video generation app violates its trademark rights. Cameo said in the complaint that OpenAI's Cameo, which allows users to create and share their virtual likenesses on Sora, is likely to cause consumer confusion and dilute its brand. Spokespeople for OpenAI did not immediately respond to a request for comment on the complaint. Cameo CEO Steven Galanis said in a statement that the company had tried to resolve the dispute "amicably" but OpenAI refused to stop using the Cameo name. Cameo, founded in 2017, allows users to hire celebrities to appear in short, personalized videos. Artificial intelligence giant OpenAI's Sora, launched as a standalone app on September 30, lets users create and share AI-generated videos. OpenAI calls Sora users' likenesses on the app "Cameos." Cameo said OpenAI chose the name "in blatant disregard for the obvious confusion it would create." Cameo also said that Sora allows users to create videos with celebrity "Cameos" including entrepreneur Mark Cuban and boxer and influencer Jake Paul, putting the AI giant in direct competition with Cameo's platform. "Users seeking a personalized celebrity video now have a choice—use Plaintiff's CAMEO service to book talent and receive an authentic, custom video prepared by that celebrity, or use Sora’s 'Cameo' service to create an extremely realistic AI-generated video featuring a celebrity’s likeness," the lawsuit said. Cameo requested an unspecified amount of monetary damages and a court order blocking OpenAI from using the "Cameo" name. (Reporting by Blake Brittain in Washington; Editing by Sam Holmes)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

SEOUL (Reuters) -SK Hynix on Wednesday said it expects the global memory chip market to experience a prolonged "super cycle", and that supply growth is likely to be limited just as demand for AI applications expands. (Reporting by Hyunjoo Jin and Heekyong Yang; Editing by Christopher Cushing)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

(Reuters) -Nvidia CEO Jensen Huang said on Tuesday he expects to make announcements with South Korean companies that will please U.S. President Donald Trump and South Korea when he visits the country this week. Huang plans to meet global leaders and top Korean executives when he attends the Asia-Pacific Economic Cooperation CEO Summit in South Korea, the U.S. artificial intelligence chip giant said earlier. "Samsung, SK, Hyundai, LG, Naver, if you just look across the entire Korean ecosystem, every one of the companies are deep friends of mine and very good partners," he told reporters on the sidelines of the Nvidia developers' conference held in Washington on Tuesday. "And so when I go hopefully we'll have some announcements that will be really, really delightful to the people of Korea and really delightful to President Trump, but I'll save it for a few more days," he said. He said Nvidia is partnering with Samsung Electronics and Hyundai "in many ways," saying the two companies are "going to invest in AI factories." Trump is due to address a summit of CEOs and meet with South Korean President Lee Jae Myung in Gyeongju on Wednesday. (Reporting by Hyunjoo Jin in Seoul and Stephen Nellis in Washington; Editing by Sonali Paul)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

SEOUL (Reuters) -SK Hynix on Wednesday said it expects the global memory chip market to experience a prolonged "super cycle", and that supply growth is likely to be limited just as demand for AI applications expands. (Reporting by Hyunjoo Jin and Heekyong Yang; Editing by Christopher Cushing)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

GYEONGJU, South Korea/Seoul (Reuters) -Amazon Web Services will invest at least $5 billion in South Korea by 2031 to build new artificial intelligence data centres, South Korea's presidential office said on Wednesday, leveraging Seoul's ambition to become an AI hub in Asia. The announcement was made during AWS Chief Executive Officer Matt Garman's meeting with South Korean President Lee Jae Myung on the sidelines of the Asia-Pacific Economic Cooperation summit.  Amazon is one of seven global firms whose executives attended the group meeting with Lee in Gyeongju, South Korea, and pledged a total of $9 billion in investments for the next five years, the presidential office said.  Amazon's investment will accelerate the growth of an ecosystem for the AI industry in South Korea, as the country aims to become one of the world's top three AI leaders, Lee said at the meeting with Garman.  "At AWS, we've invested and committed to investment of an additional $40 billion across 14 non-U.S. APEC countries and economies between now and 2028," Garman said. "And, that $40 billion actually drives an additional $45 billion in U.S. GDP and downstream benefit, benefiting all of the APEC economy," he said at a business event on the sidelines of the summit. AWS plans to build the new data centres on the outskirts of Seoul. AWS unveiled in June a $4 billion investment plan in South Korea as part of SK Group's investments to build the country's biggest data centre in the southeastern city of Ulsan. It has also announced investments in other countries, including Japan, Australia and Singapore. Earlier this month, South Korea's presidential office said OpenAI planned to set up joint ventures with Samsung and SK to build two data centres, a Korean-style Stargate, with an initial capacity of 20 megawatts. OpenAI said South Korea has one of the largest numbers of paying ChatGPT subscribers, second only to the United States. OpenAI also signed initial deals to source memory chips for its data centres from Samsung Electronics and SK Hynix. (Reporting by Heejin Kim, Jihoon Lee and Hyunjoo Jin; Editing by Himani Sarkar)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Heekyong Yang, Hyunjoo Jin and Joyce Lee SEOUL (Reuters) -South Korea's SK Hynix has sold out all its chip production for next year and plans to sharply boost investments, expecting an extended chip "super cycle" spurred by the AI boom, it said on Wednesday after reporting a record quarterly profit. The company, which supplies artificial intelligence chip leader Nvidia, said supply growth for memory chips is likely to be limited just as demand is surging, with customers vying to expand investments in AI data centre infrastructure. SK Hynix shares rose as much as 6% after the record profit, outpacing a 1.5% rise in the benchmark KOSPI. A growing number of customers are scrambling to lock in supplies in response to a chip shortage, placing orders for next year in advance and pushing for long-term contracts, the company said. At the same time, a shift in production to high-bandwidth memory (HBM) has strained output of conventional memory, such as DRAM, and "total output will inevitably remain limited," said Kim Kyu-hyun, SK Hynix's head of DRAM marketing. "This structural constraint on DRAM supply is expected to support the current prolonged memory super cycle, as supply growth lags behind accelerating demand," he said. SK Hynix reported a record 11.4 trillion won ($8.02 billion) operating profit for the July-September period, up 62% from a year earlier and in line with a forecast by LSEG SmartEstimate. Quarterly revenue rose 39% in the third quarter to 24.4 trillion won. DEMAND GROWTH OUTSTRIPS SUPPLY SK Hynix has already secured full customer demand for its entire DRAM, HBM, and NAND production for next year and plans to start shipping the next-generation HBM4 chips in the fourth quarter of this year. When asked about its HBM pricing for next year, the company said it would be able to maintain profitability on its HBM products. The chip maker expects DRAM shipments to grow by more than 20% year-on-year in 2026. "DRAM inventories have dropped to extremely low levels, and for DDR5 products, newly produced chips must be shipped to customers immediately to meet demand," said Kim. While thriving on a strong market for conventional memory, the company wants to maintain its edge in advanced AI chips supplied to Nvidia, in the face of mounting competition from rivals such as Samsung Electronics and Micron. SK Hynix said it had completed discussions with key customers regarding HBM supply for 2026, and plans a full-scale sales expansion of HBM4 products next year. SK Hynix said the HBM market is expected to post annual growth of over 30% over the next five years, supported by expanding investments from big tech and AI companies, even under a conservative outlook. The company added that its recent letter of intent with ChatGPT maker OpenAI for HBM supply underscores strong AI-driven demand and the critical importance of securing HBM products. As confidence grows in the earnings potential of the AI market, global AI companies are ramping up their investment spending, Kim said, driving rapid growth in demand for a wide range of memory products including HBM, DDR5, and enterprise SSDs. To meet this demand, a rise in capital expenditure across the memory industry is inevitable, and SK Hynix said it expects its own spending to grow substantially from this year’s level, without specifying a figure. Buoyed by its advantage in advanced chips, SK Hynix's shares have surged 200% so far this year, outpacing an 87% gain in Samsung's shares and a 67% gain in the benchmark KOSPI. ($1 = 1,421.4800 won) (Reporting by Heekyong Yang, Hyunjoo Jin and Joyce Lee; Editing by Jamie Freed, Sonali Paul and Christian Schmollinger)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Wen-Yee Lee TAIPEI (Reuters) -Foxconn, the world's largest electronics maker and Nvidia's key AI server maker, said on Tuesday it will deploy humanoid robots at its Houston plant that produces AI servers for Nvidia. Reuters first reported in June that Foxconn and Nvidia were in talks to use humanoid robots at the Houston factory, targeting the first quarter of 2026. "The factory will also be among the first to deploy humanoid robots powered by the NVIDIA Isaac GR00T N model on its production lines, as Foxconn and Nvidia aim to build a world-leading benchmark AI smart factory," the company said in a statement released during Nvidia’s developers' conference in Washington, D.C. Foxconn, formally known as Hon Hai Precision Industry Co, said it will also continue scaling up AI server production in Texas, Wisconsin and California to meet rising demand. “Our team is bringing the most advanced AI data center solutions to the United States, which will help our leading customers stay ahead in the AI race,” Foxconn Chairman Young Liu said in the release. (Reporting by Wen-Yee Lee; Editing by Sonali Paul)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

GYEONGJU, South Korea (Reuters) -Amazon Web Services will invest at least $5 billion in South Korea by 2031 to build new artificial intelligence data centres in the Asian country, South Korea's presidential office said on Wednesday. (Reporting by Heejin Kim; Editing by Himani Sarkar)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

GYEONGJU, South Korea (Reuters) -Amazon Web Services will invest at least $5 billion in South Korea by 2031 to build new artificial intelligence data centres in the Asian country, South Korea's presidential office said on Wednesday. (Reporting by Heejin Kim; Editing by Himani Sarkar)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Heekyong Yang, Hyunjoo Jin and Joyce Lee SEOUL (Reuters) -South Korea's SK Hynix has sold out all its chip production for next year and plans to sharply boost investments, expecting an extended chip "super cycle" spurred by the AI boom, it said on Wednesday after reporting a record quarterly profit. The company, which supplies artificial intelligence chip leader Nvidia, said supply growth for memory chips is likely to be limited just as demand is surging, with customers vying to expand investments in AI data centre infrastructure. SK Hynix shares rose as much as 6% after the record profit, outpacing a 1.5% rise in the benchmark KOSPI. A growing number of customers are scrambling to lock in supplies in response to a chip shortage, placing orders for next year in advance and pushing for long-term contracts, the company said. At the same time, a shift in production to high-bandwidth memory (HBM) has strained output of conventional memory, such as DRAM, and "total output will inevitably remain limited," said Kim Kyu-hyun, SK Hynix's head of DRAM marketing. "This structural constraint on DRAM supply is expected to support the current prolonged memory supercycle, as supply growth lags behind accelerating demand," he said. SK Hynix reported an 11.4 trillion won ($8.02 billion) operating profit for the July-September period, up 62% from a year earlier and in line with a forecast by LSEG SmartEstimate. Quarterly revenue rose 39% in the third quarter to 24.4 trillion won. DEMAND GROWTH OUTSTRIPS SUPPLY SK Hynix has already secured full customer demand for its entire DRAM and NAND production for next year and plans to start shipping the next-generation HBM4 chips in the fourth quarter of this year. The chip maker expects DRAM shipments to grow by more than 20% year-on-year in 2026. "DRAM inventories have dropped to extremely low levels, and for DDR5 products, newly produced chips must be shipped to customers immediately to meet demand," said Kim. While thriving on a strong market for conventional memory, the company wants to maintain its edge in advanced AI chips supplied to Nvidia, in the face of mounting competition from rivals such as Samsung Electronics and Micron. SK Hynix said it had completed discussions with key customers regarding HBM supply for 2026, and plans a full-scale sales expansion of HBM4 products next year. As confidence grows in the earnings potential of the AI market, global AI companies are ramping up their investment spending, Kim said, driving rapid growth in demand for a wide range of memory products including HBM, DDR5, and enterprise SSDs. To meet this demand, a rise in capital expenditure across the memory industry is inevitable, and SK Hynix said it expects its own spending to grow substantially from this year’s level, without specifying a figure. Buoyed by its advantage in advanced chips, SK Hynix's shares have surged 200% so far this year, outpacing an 87% gain in Samsung's shares and a 67% gain in the benchmark KOSPI. ($1 = 1,421.4800 won) (Reporting by Heekyong Yang, Hyunjoo Jin and Joyce Lee; Editing by Jamie Freed and Sonali Paul)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)