Top related persons:
Top related locs:
Top related orgs:

Search resuls for: "SemiAnalysis"


25 mentions found


AdvertisementTSMC's Phoenix chip factories likely won't eliminate US dependence on tech from Taiwan. Experts told BI that TSMC's most advanced chips will likely continue to be produced in Taiwan. TSMC produces an estimated 90% of the world's advanced chips, which power everything from iPhones to cars. While TSMC's Phoenix factories are expected to boost semiconductor chip production in the US, the company isn't making its most advanced chips stateside, industry experts told Business Insider. For example, chips produced using 4 nanometer (nm) technology are expected to be made in the first Phoenix factory, while TSMC's Taiwan factories are already producing chips using 3nm technology.
Persons: Biden, TSMC, Jeff Koch, SemiAnalysis, — Koch, Stephen Ezell, it'll, Chris Miller, Miller, Koch, Trump, William Alan Reinsch Organizations: Commerce Department, Information Technology, Innovation Foundation, of Commerce, Arizona, Nvidia, Apple, Qualcomm, AMD, American Enterprise Institute, Center for Strategic, International Studies, Biden Locations: Taiwan, Arizona, Phoenix, TSMC's, United States, China
AdvertisementNvidia's Blackwell chip presents cooling challenges for customers. First, there were chip design issues, which CEO Jensen Huang has since said are fully resolved. "I think the overheating issues have been present for months and they have largely been addressed," Patel told Insider. AdvertisementIn addition to engineering and operational challenges, liquid cooling at scale brings with it a list of environmental concerns. Despite the hard work and environmental strain, of converting to liquid cooling, the incentives are strong.
Persons: Nvidia's Blackwell, Blackwell, Jensen Huang, Dylan Patel, Patel, Meta, Eugene Kim, Semianalysis, Huang Organizations: Nvidia, Blackwell, Semianalysis, BI, Amazon Locations: Semianalysis, Blackwell
AdvertisementThe Biden administration is trying to finalize CHIPS Act agreements before Donald Trump takes office. Locking in agreements could make it more difficult for the Trump administration if it decides to change course. The Biden administration is running out of time to finalize agreements that would secure billions in funding for US chipmakers. The Biden administration is working to finalize agreements before president-elect Donald Trump — who has criticized the CHIPS Act — takes office in January. However, she did not answer questions about whether Trump's approach to CHIPS Act funding could differ from that of the Biden Administration.
Persons: Biden, Donald Trump, Trump, Joe Biden, hasn't, Donald Trump —, Gina Raimondo, podcaster Joe Rogan, chipmakers, Mike Johnson, Karoline Leavitt, Vance, Jeff Koch, SemiAnalysis, Koch, Pat Gelsinger Organizations: US, Bloomberg, Commerce Department, BI, CNBC, Reuters, Commerce, Biden, Intel, The New York Times Locations: Taiwan
AdvertisementThe tariffs Donald Trump proposed while campaigning could help create US semiconductor jobs. To be sure, the impacts of a tariff policy on chips could be mixed. He said that the federal government should have implemented tariffs instead to motivate global chipmakers to build more factories in the US, which would then create jobs. AdvertisementTo be sure, the impacts of a tariff policy on chips could be mixed. In August, the White House said that investments supported by CHIPS Act funding would help create 115,000 construction and manufacturing jobs.
Persons: Donald Trump, podcaster Joe Rogan, Trump, Jeff Koch, SemiAnalysis, Patrick Moorhead, Jeff Ferry, Ferry, Stephen Ezell, Ezell Organizations: US, Industry, Biden, Samsung, Coalition for, Prosperous, Semiconductor Industry Association —, Intel, Information Technology, Innovation Foundation, Apple Locations: Taiwan, China, South Korea, Prosperous America, Arizona, India, United States
US efforts to produce semiconductor chips will continue regardless of who wins the election. AdvertisementNo matter who wins the presidential election this November, President Joe Biden can rest easy knowing one thing: The US's chip manufacturing push isn't going anywhere. The US has seen its share of overall chip production fall from 37% of the world's supply in 1990 to 12%. While many factories remain under construction, the federal funding has already helped boost US chip production. According to a report published last year by the trade and lobbying group Semiconductor Industry Association and Oxford Economics, the US semiconductor industry will face a shortage of 67,000 workers by 2030, including technicians, computer scientists, and engineers.
Persons: Harris, Trump, , Joe Biden, Biden, Kamala Harris, Donald Trump, Stephen Ezell, Anna Kelly, Kelly, Arizonans, Chris Miller, Mark Muro, Dylan Patel, SemiAnalysis, Ezell, Muro Organizations: Service, Trump, Information Technology, Innovation Foundation, ITIF's Center for Life Sciences, Biden, Republican, Intel, American Enterprise Institute, Brookings Metro, Apple, Nvidia, Democratic, Semiconductor Industry Association, Oxford Economics Locations: Arizona, USA, China, Taiwan
AdvertisementThe US's efforts to produce more semiconductor chips have encountered some challenges over the past few years, but the tide may be turning. In 2022, President Joe Biden signed the CHIPS Act into law, which included $39 billion in manufacturing incentives for chip production in the US. Last year, TSMC announced that the official opening of its first Arizona fab would be pushed back from 2024 to 2025. TSMC's production of Apple chips is good news for Americans hoping to land jobs in the industry, Patel said. He said he thinks TSMC's production of Apple chips would bode well for the future.
Persons: TSMC, Biden, It's, , Tim Culpan, Culpan, Tim Cook, Dylan Patel, SemiAnalysis, Mark Muro, Apple, Harris, Joe Biden, Arizona hasn't, it's, Morris Chang, Patel, Muro, bode Organizations: Apple, Service, Brookings Institution, Biden, Management, Workers, Brookings Institute, Semiconductor Industry Association Locations: Arizona, Taiwan, Phoenix
Liang told Business Insider he expects 90% of AI computing workloads will be in inference in the not-too-distant future. AdvertisementThat's why several startups are charging aggressively into the inference market — emphasizing where they might outperform the goliath in the space. Speed is an important factor when multiple AI models talk to each other and waiting for an answer can dampen the magic of generative AI. The number of tokens per second that can be consumed (when a prompt goes in) and generated (when a response comes out) is a common metric for AI computing speed. AdvertisementCerebras's AI chip is roughly size of a dinner plate.
Persons: , Rodrigo Liang, Colleen Kress, workloads, Liang, Andrew Feldman, Bernstein, Colette Kress, Kress, Nvidia's, SambaNova, Anthropic's Claude, Nvidia isn't, Jensen, Dylan Patel, Patel Organizations: Service, Business, SambaNova Systems, Nvidia, ARM, AMD, o1, BI Locations: Artificialanalysis.ai, TCO
Qualcomm could sell parts of Intel to other buyers to get any deal done, the newspaper added. Intel mostly sells chips for PCs and data center servers, although Nvidia's GPUs have made serious inroads into this lucrative data center business. An "odd" fitSemianalysis chief analyst Dylan Patel said Qualcomm and Intel would be an "odd" fit. "Furthermore, Qualcomm has no ability to turn around the data center business, which is the most important one," he continued. The company's revenue has declined in recent years while rivals like Nvidia, TSMC, and Broadcom are riding the artificial intelligence wave to new heights.
Persons: , Apple iPhones, Dylan Patel, Patel, Pat Gelsinger, Gelsinger Organizations: Service, Qualcomm, Intel, Wall Street, Business, Nvidia, Apple, Intel's Foundry, TSMC, Broadcom, Amazon Web Services, Gaudi, AMD, UXL, Google
When Nvidia CEO Jensen Huang spoke to Goldman Sachs CEO David Solomon on Wednesday, the banker raised this relationship — and how precarious it is due to rising tension in the Taiwan Strait. This story is available exclusively to Business Insider subscribers. Last year, Huang said he felt "perfectly safe" relying on manufacturing in Taiwan. Though Samsung has some of the same capabilities required to produce the most advanced chips, TSMC is the clear leader. Huang went on to praise TSMC for scaling up to meet the demand that created Nvidia's "hockey stick" earnings in recent years.
Persons: , Jensen Huang, Goldman Sachs, David Solomon, Solomon, Huang, Dylan Patel, Chris Miller, TSMC, Patel Organizations: Service, Nvidia, Wednesday, Business, Samsung, Intel, SEC, Blackwell Locations: Taiwan, Asia, China, Pacific, TSMC, South, Arizona
Read previewJensen Huang faces a monumental task as he takes center stage at a Goldman Sachs event on Wednesday: getting AI investors to stop panicking. However, there are signs suggesting that investors are getting pretty angsty over the future of Nvidia and AI more broadly. The markets are in a panic over AINvidia has been a powerhouse at the center of the generative AI boom. First, Nvidia's 122% year-on-year revenue growth in the last quarter was smaller than the 262% growth it recorded in the previous quarter. "I do not believe that Jensen is able to say anything that will quell investors' nerves," he said.
Persons: , Jensen Huang, Goldman Sachs, Chris Beauchamp, Jeff Chiu, Beauchamp, Bernstein, Emma Cosgrove, Huang, Blackwell, Alvin Nguyen, Forrester, Jensen Organizations: Service, Nvidia, Technology Conference, Business, Google, Meta, Microsoft, Santa, Big Tech, Blackwell Locations: San Francisco, Santa Clara
Its next-generation Blackwell chip faces delays, raising questions about supply-chain impacts. As Nvidia prepares to report its second-quarter earnings on Wednesday, some of those questions are still relevant, while some have faded away. Efforts by other companies to build competitive AI chips haven't reached more than a simmer, and the rollout of Nvidia's next-generation Blackwell chip is said to be delayed. The analysts also wrote that revenue from existing products, specifically the H200 and H20 AI chips, could offset any delayed revenue from the Blackwell chip. AdvertisementNvidia's chips are still a must for the most complex AI-computing jobs, as Amazon still purchases chips from Nvidia.
Persons: Blackwell, , Nvidia's, ChatGPT, China . Blackwell, weren't, Dylan Patel, Morgan Stanley, Dan Morgan, Meta, Mark Zuckerberg, we're, Zuckerberg, Lisa Su, Synovus, Morgan, Jensen Huang Organizations: Nvidia, Service, North Star, Google, Microsoft, JPMorgan, Amazon, AMD, Reuters Locations: China, Synovus
download the appSign up to get the inside scoop on today’s biggest stories in markets, tech, and business — delivered daily. SemiAnalysis gets deep into the weeds on graphics processing units, from the base of the supply chain to the end-users building the AI models that require so many specialized chips. AdvertisementThe team consists of people across the entire supply chain, geographically and technically. So we have the entire view of the supply chain, from manufacturing, up to models. The other place I like to look a lot is random suppliers that are very small in niche parts of the supply chain, or suppliers of suppliers.
Persons: , Jensen Huang isn't, Dylan Patel, SemiAnalysis, Sam Altman, Patel, Sundar, Altman, Blackwell, Nvidia's, They've, they're Organizations: Service, intel, Google, Business, BI, Nvidia, Microsoft, AMD, Meta Locations: US, Japan, Singapore, Taiwan, France
The rollout of Nvidia's next-generation of AI chips, Blackwell, is going to be delayed according to media and industry analyst reports. Smaller cloud firms that have built their businesses on Nvidia chips could be affected by the delay too. Each new generation of AI chips promises step changes in computing capacity, speed and efficiency, allowing for faster computations and larger models. Taiwan Semiconductor Manufacturing Company, Nvidia's manufacturing partner, is using a relatively new process to manufacture the Blackwell chips. Some Blackwell chips have begun shipping to customers and production will still increase in the second half of the year.
Persons: , Blackwell, Semianalysis, Hopper, Jensen Huang, Huang, there's, Nvidia's Blackwell, Bernstein Organizations: Service, Blackwell, Business, Nvidia, Meta, Microsoft, Google, AMD, Taiwan Semiconductor Manufacturing Company, Reuters, Bank of America Locations: Semianalysis
Nvidia is working on a version of its new flagship AI chips for the China market that would be compatible with current U.S. export controls, four sources familiar with the matter said. The AI chip giant in March unveiled its "Blackwell" chip series, which is due to be mass-produced later in the year. A version of a chip from Nvidia's Blackwell series for the Chinese market would boost the U.S. firm's efforts to fend off those challenges. Expectations are high that the U.S. will continue to keep up the pressure on semiconductor-related export controls. The U.S. wants the Netherlands and Japan to further restrict chipmaking equipment to China, sources have said.
Persons: Blackwell, Inspur, Nvidia's Blackwell, Biden Organizations: MWC Shanghai, Nvidia, Reuters, Huawei, Bloomberg News Locations: Shanghai, China, Washington, U.S, Netherlands, Japan
Share Share Article via Facebook Share Article via Twitter Share Article via LinkedIn Share Article via EmailMicron's profit margin will rise given its growing AI memory business: AnalystDylan Patel from SemiAnalysis talks about Micron's business outlook as the market is starting the one of the largest memory cycles in history.
Persons: Dylan Patel, SemiAnalysis
Last year, Taiwan Semiconductor Manufacturing Company reported its first profit decline in four years. The company said engineering wafer production began at the factory in April, an important step toward the eventual chip production. Last July, TSMC announced that chip production for the first factory would be postponed from 2024 to 2025. Barring further setbacks, TSMC's update could mean the first factory will begin production of chips in 2025. This funding could be particularly important for TSMC, given the cost of factory construction and chip manufacturing can differ between the US and Taiwan.
Persons: , TSMC, Biden, Joe Biden, — TSMC, Dylan Patel, SemiAnalysis, that's, Morris Chang, C.C, Wei Organizations: Service, Taiwan Semiconductor Manufacturing Company, Business, Biden, Financial Times Locations: Taiwan, Arizona, Phoenix, Japan, Germany
The demand for Nvidia's AI chips could bolster job growth across the semiconductor industry. It could also help bring more semiconductor chip manufacturing stateside and reduce the US's reliance on Taiwan — which remains vulnerable to Chinese invasion that would wreak havoc on the global economy. AdvertisementEven without the AI boom, the semiconductor industry was already poised for big job gains in the near future. Of the 115,000 US new semiconductor jobs the Semiconductor Industry Association is projecting by 2030, it said roughly 67,000 of these positions risk going unfilled given current college degree completion rates. In an effort to prevent a worker shortage, community colleges and universities across the country have partnered with semiconductor companies.
Persons: , Pat Gelsinger, It's, who've, TSMC, Syed Alam, Ed Kaste, GlobalFoundries, Mark Muro, Muro, Jensen Huang, hasn't, Dylan Patel, SemiAnalysis, Patel, Accenture's Alam, Alam Organizations: Nvidia, TSMC, Service, Deloitte, Semiconductor Industry Association, Accenture, Meta, Intel, AMD, Brookings Institution, Samsung, Google, IBM, Lam Research, Materials, KLA Corporation Locations: Taiwan, Arizona
Nvidia is in talks with tech giants like OpenAI and Google to build custom AI chips. It's held talks with leaders from Meta, Microsoft, Google, and OpenAI to build custom chips for data centers, two sources told Reuters. AdvertisementAny move by Nvidia into the custom chip market bodes ill for other manufacturers. Nvidia isn't just trying to work with major tech companies. The company is also turning to its own AI to produce its own AI chips faster.
Persons: , It's, Nvidia's, OpenAI, Greg Reichow, Dylan Patel, Nvidia isn't, Sam Altman, Jensen Huang, Aaron Mok Organizations: Nvidia, Google, Broadcom, Reuters, Service, Meta, Microsoft, Business, Marvell Technology, Eclipse Ventures, OpenAI, United Arab Emirates
The dominant global designer and supplier of AI chips aims to capture a portion of an exploding market for custom AI chips and to protect itself from the growing number of companies interested in finding alternatives to its products. Nvidia officials have met with representatives from Amazon.com , Meta, Microsoft, Google and OpenAI to discuss making custom chips for them, according to two sources familiar with the meetings. $30 billion marketAccording to estimates from research firm 650 Group's Alan Weckel, the data center custom chip market will grow to as much as $10 billion this year, and double that in 2025. The broader custom chip market was worth roughly $30 billion in 2023, which amounts to roughly 5% of annual global chip sales, according to Needham analyst Charles Shi. "With Broadcom's custom silicon business touching $10 billion, and Marvell's around $2 billion, this is a real threat," said Dylan Patel, founder of the silicon research group SemiAnalysis.
Persons: OpenAI, Greg Reichow, Meta, Dina McKinney, Alan Weckel, Charles Shi, Dylan Patel Organizations: Nvidia, Microsoft, Broadcom, Marvell Technology, Eclipse Ventures, Amazon.com, Meta, Google, Reuters, Devices, Marvell, Taiwan Semiconductor Manufacturing Locations: Krakow, Poland, Santa Clara , California
The California-based AI chip giant had been expected to launch the new products as early as Nov. 16, chip industry newsletter SemiAnalysis reported this month. However, the H20 launch has now been pushed back until the first quarter of next year, the sources said, with one saying they were advised it could take place in February or March. In addition to the H20, Nvidia has been planning two other chips to comply with new U.S. export rules - the L20 and L2. The sources said the L20 was not facing delays and would launch according to its original schedule. Chinese internet giant Baidu (9888.HK) placed a sizeable order for Huawei AI chips this year, Reuters reported this month citing sources.
Persons: Dado Ruvic, Baidu, Fanny Potkin, Yelin, Brenda Goh, Jamie Freed Organizations: NVIDIA, REUTERS, Rights, Nvidia, Huawei, U.S, Baidu, HK, Reuters, Thomson Locations: Rights SINGAPORE, BEIJING, China, The California, U.S, Singapore, Yelin Mo, Beijing, Shanghai
Microsoft emerges as big winner from OpenAI turmoil
  + stars: | 2023-11-20 | by ( Aditya Soni | ) www.reuters.com   time to read: +3 min
[1/2] Microsoft Chief Executive Officer (CEO) Satya Narayana Nadella speaks at a live Microsoft event in the Manhattan borough of New York City, October 26, 2016. The turmoil at OpenAI since Friday had raised fears about the fallout for Microsoft, which has pumped in billions of dollars and uses the pioneer's technology for most of its AI offerings. That was nearly equal to the valuation OpenAI commanded in its last fundraise. We view Microsoft now even in a stronger position with Altman and Brockman at Microsoft running AI. "The OpenAI for-profit subsidiary was about to conduct a secondary at a $80 billion+ valuation.
Persons: Satya Narayana Nadella, Lucas Jackson, Sam Altman, Altman, Dan Ives, Brockman, OpenAI, Greg Brockman, Szymon Sidor, SemiAnalysis, Mira Murati, Aditya Soni, Akash Sriram, Sriraj Organizations: Microsoft, REUTERS, Google, Wedbush Securities, Thomson Locations: Manhattan, New York City, OpenAI, Bengaluru
That was nearly equal to the valuation OpenAI commanded in its last fundraise. "If Microsoft lost Altman, he could have gone to Amazon, Google, Apple, or a host of other tech companies," said Wedbush Securities analyst Dan Ives. We view Microsoft now even in a stronger position with Altman and Brockman at Microsoft running AI. Altman will lead a new research team at the software giant following his surprise ouster by OpenAI that shocked the tech industry. "The OpenAI for-profit subsidiary was about to conduct a secondary at a $80 billion+ valuation.
Persons: Satya Narayana Nadella, Lucas Jackson, Sam Altman, Altman, Dan Ives, Brockman, OpenAI, Greg Brockman, Szymon Sidor, SemiAnalysis, Mira Murati, Aditya Soni, Akash Sriram, Sriraj Organizations: Microsoft, REUTERS, Google, Wedbush Securities, Thomson Locations: Manhattan, New York City, OpenAI, Bengaluru
Tencent has enough Nvidia chips to continue development of its "Hunyuan" AI model "for at least a couple more generations", so the curbs will not affect near-term AI capability, Lau said. "We will have to figure out ways to make the usage of our AI chips more efficient," he said. "And we will also try to look for domestic sources for these training chips." "We feel that the chip ban does actually affect our ability to resell (use of) these AI chips through our cloud services," he said. Nvidia plans to market new China-bound AI chips, with an announcement on Nov. 16 at the earliest, industry newsletter SemiAnalysis reported last week.
Persons: Aly, Martin Lau, Lau, We'll, SemiAnalysis, Josh Ye, Jane Merriman, Christopher Cushing Organizations: Artificial Intelligence, REUTERS, Tencent Holdings, HK, Nvidia, Huawei Technologies, Reuters, Baidu, Huawei, Thomson Locations: Shanghai, China, HONG KONG, U.S
A smartphone with a displayed NVIDIA logo is placed on a computer motherboard in this illustration taken March 6, 2023. The chip industry newsletter SemiAnalysisas reported earlier that Nvidia plans to release new artificial intelligence chips aimed at the Chinese market less than a month after the U.S. tightened rules on selling high-end AI chips to China. The new rules put a cap on how much computing power a chip can pack into a small size. SemiAnalysis said the new Nvidia chips are called the HGX H20, L20 PCIe and L2 PCIe and the company could announce them on Nov. 16. "It is not difficult to imagine that as long as Washington remains committed to 'choking' China, the game of 'catch me if you can' will continue indefinitely," the newspaper said.
Persons: Dado Ruvic, SemiAnalysis, Robert Birsel Organizations: NVIDIA, REUTERS, Rights, Nvidia, Global Times, U.S, Shanghai, Thomson, Reuters Locations: China, U.S, Washington
Tyrone Siu | ReutersBEIJING — U.S. chipmaking giant Nvidia has reportedly found a way to sell high-end chips to Chinese companies — while remaining compliant with U.S. rules aimed at curbing China's access to the tech. Nvidia is set to deliver three new chips to domestic manufacturers in the coming days, Chinese financial media Cailian Press said Thursday, citing sources. The chips — called HGX H20, L20 PCle and L2 PCle — are based on Nvidia's H100 chip, the report said. Nvidia said in a September 2022 filing the U.S. government would still allow it to develop the H100 in China. In the near term, Chinese manufacturers have no better option and they will continue to buy Nvidia's chips, while searching for replacements.
Persons: Tyrone Siu, Bo Du, Du, SemiAnalysis, Nomura, Orin Organizations: Nvidia Corporation, Reuters, Reuters BEIJING —, Nvidia, Cailian Press, WestSummit Capital Management Companies, WestSummit Capital Management, CNBC, Times, U.S . Department of Commerce, Bureau of Industry, Security Locations: Taipei, Taiwan, Reuters BEIJING, China, U.S
Total: 25