Latest news with #Huang
Time of India
an hour ago
- Business
- Time of India
Nvidia to launch cheaper Blackwell AI chip for China after US export curbs, sources say
Nvidia will launch a new artificial intelligence chipset for China at a significantly lower price than its recently restricted H20 model and plans to start mass production as early as June, sources familiar with the matter said. The GPU or graphics processing unit will be part of Nvidia's latest generation Blackwell-architecture AI processors and is expected to be priced between $6,500 and $8,000, well below the $10,000-$12,000 the H20 sold for, according to two of the sources. The lower price reflects its weaker specifications and simpler manufacturing requirements. It will be based on Nvidia's RTX Pro 6000D, a server-class graphics processor and will use conventional GDDR7 memory instead of more advanced high bandwidth memory, the two sources said. They added it would not use Taiwan Semiconductor Manufacturing Co's advanced Chip-on-Wafer-on-Substrate (CoWoS) packaging technology. The new chip's price, specifications and production timing have not previously been reported. The three sources Reuters spoke to for this article declined to be identified as they were not authorised to speak to media. An Nvidia spokesperson said the company was still evaluating its "limited" options. "Until we settle on a new product design and receive approval from the U.S. government, we are effectively foreclosed from China's $50 billion data center market." TSMC declined to comment. Market share plunge China remains a huge market for Nvidia, accounting for 13% of its sales in the past financial year. It's the third time that Nvidia has had to tailor a GPU for the world's second-largest economy after restrictions from U.S. authorities who are keen to stymie Chinese technological development. After the U.S. effectively banned the H20 in April, Nvidia initially considered developing a downgraded version of the H20 for China, sources have said, but that plan didn't work out. Nvidia CEO Jensen Huang said last week the company's older Hopper architecture - which the H20 uses - can no longer accommodate further modifications under current U.S. export restrictions. Reuters was unable to determine the product's final name. Chinese brokerage GF Securities said in a note published on Tuesday that the new GPU would likely be called the 6000D or the B40, though it did not disclose pricing or cite sources for the information. According to two of the sources, Nvidia is also developing another Blackwell-architecture chip for China that is set to begin production as early as September. Reuters was not immediately able to confirm specifications of that variant. Nvidia's market share in China has plummeted from 95% before 2022, when U.S. export curbs that impacted its products began, to 50% currently, Huang told reporters in Taipei this week. Its main competitor is Huawei which produces the Ascend 910B chip. Huang also warned that if U.S. export curbs continue, more Chinese customers will buy Huawei's chips. The H20 ban forced Nvidia to write off $5.5 billion in inventory and Huang told the Stratechery podcast on Monday that the company also had to walk away from $15 billion in sales. The latest export restrictions introduced new limits on GPU memory bandwidth - a crucial metric measuring data transmission speeds between the main processor and memory chips. This capability is particularly important for AI workloads that require extensive data processing. Investment bank Jefferies estimates that the new regulations cap memory bandwidth at 1.7-1.8 terabytes per second. That compares with the 4 terabytes per second that the H20 is capable of. GF Securities forecast the new GPU will achieve approximately 1.7 terabytes per second using GDDR7 memory technology, just within the export control limits.
Business Recorder
3 hours ago
- Science
- Business Recorder
Astronomers spot galaxy shaped like the Milky Way but is far more massive
WASHINGTON: Astronomers have observed a galaxy dating to an earlier epoch in the universe's history that surprisingly is shaped much like our Milky Way - a spiral structure with a straight bar of stars and gas running through its center - but far more massive, offering new insight into galactic formation. The distant galaxy, called J0107a, was observed as it appeared 11.1 billion years ago, when the universe was about a fifth of its current age. The researchers used data from the Chile-based Atacama Large Millimeter/submillimeter Array (ALMA) and NASA's James Webb Space Telescope to study the galaxy. They determined that the galaxy's mass, including its stars and gas, was more than 10 times greater than that of the Milky Way, and it was forming stars at an annual rate approximately 300 times greater. J0107a was more compact than the Milky Way, however. 'The galaxy is a monster galaxy with a high star formation rate and plenty of gas, much more than present-day galaxies,' said astronomer Shuo Huang of the National Astronomical Observatory of Japan, lead author of the study published this week in the journal Nature. 'This discovery,' said study co-author Toshiki Saito, an astronomer at Shizuoka University in Japan, 'raises the important question: How did such a massive galaxy form in such an early universe?' While a few galaxies that are undergoing star formation at a similar rate to J0107a exist in today's universe, almost all of them are ones that are in the process of a galactic merger or collision. There was no sign of such circumstances involving this galaxy. J0107a and the Milky Way have some commonalities. 'They are similarly huge and possess a similar barred structure. However, the Milky Way had plenty of time to form its huge structures, while J0107a didn't,' Saito said. In the first few billion years after the Big Bang event 13.8 billion years ago that initiated the universe, galaxies were turbulent entities and were much richer in gas than those existing currently - factors that fostered extreme bursts of star formation. While galaxies with highly organized structures like the barred spiral shape of the Milky Way are common now, that was not the case 11.1 billion years ago. 'Compared to other monster galaxies in the distant universe (dating to an earlier cosmic epoch) whose shapes are usually disturbed or irregular, it is unexpected that J0107a looks very similar to present-day spiral galaxies,' Huang said. 'Theories about the formation of present-day galactic structures may need to be revised,' Huang added. The Webb telescope, as it peers across vast distances back to the early universe, has found that galaxies with a spiral shape appeared much earlier than previously known. J0107a is now one of the earliest-known examples of a barred spiral galaxy. About two thirds of spiral galaxies observed in the universe today possess a bar structure. The bar is thought to serve as a form of stellar nursery, bringing gas inward from the galaxy's spiral arms. Some of the gas forms what are called molecular clouds. Gravity causes the contraction of these clouds, with small centers taking shape that heat up and become new stars. The bar that is part of J0107a measures about 50,000 light years in length, Huang said. A light-year is the distance light travels in a year, 5.9 trillion miles (9.5 trillion km). The Webb telescope 'has been studying the morphology of early massive galaxies intensely recently. However, their dynamics are still poorly understood,' Saito said.
Straits Times
4 hours ago
- Business
- Straits Times
Nvidia to launch cheaper Blackwell AI chip for China after US export curbs, sources say
It's the third time that Nvidia has had to tailor an AI chip for China after restrictions from US authorities. PHOTO: REUTERS Nvidia to launch cheaper Blackwell AI chip for China after US export curbs, sources say BEIJING/TAIPEI - Nvidia will launch a new artificial intelligence chipset for China at a significantly lower price than its recently restricted H20 model and plans to start mass production as early as June, sources familiar with the matter said. The GPU or graphics processing unit will be part of Nvidia's latest generation Blackwell-architecture AI processors and is expected to be priced between US$6,500 (S$8,348) and US$8,000, well below the US$10,000-US$12,000 the H20 sold for, according to two of the sources. The lower price reflects its weaker specifications and simpler manufacturing requirements. It will be based on Nvidia's RTX Pro 6000D, a server-class graphics processor and will use conventional GDDR7 memory instead of more advanced high bandwidth memory, the two sources said. They added it would not use Taiwan Semiconductor Manufacturing Co's advanced Chip-on-Wafer-on-Substrate (CoWoS) packaging technology. An Nvidia spokesperson said the company was still evaluating its 'limited' options. 'Until we settle on a new product design and receive approval from the US government, we are effectively foreclosed from China's US$50 billion data center market.' China remains a huge market for Nvidia, accounting for 13 per cent of its sales in the past financial year. It's the third time that Nvidia has had to tailor a GPU for the world's second-largest economy after restrictions from US authorities who are keen to stymie Chinese technological development. After the US effectively banned the H20 in April, Nvidia initially considered developing a downgraded version of the H20 for China, sources have said, but that plan didn't work out. Nvidia chief executive officer Jensen Huang said last week the company's older Hopper architecture - which the H20 uses - can no longer accommodate further modifications under current US export restrictions. Reuters was unable to determine the product's final name. Chinese brokerage GF Securities said in a note published on May 20 that the new GPU would likely be called the 6000D or the B40, though it did not disclose pricing or cite sources for the information. According to two of the sources, Nvidia is also developing another Blackwell-architecture chip for China that is set to begin production as early as September. Reuters was not immediately able to confirm specifications of that variant. Nvidia's market share in China has plummeted from 95 per cent before 2022, when US export curbs that impacted its products began, to 50 per cent currently, Mr Huang told reporters in Taipei this week. Its main competitor is Huawei which produces the Ascend 910B chip. Mr Huang also warned that if US export curbs continue, more Chinese customers will buy Huawei's chips. The H20 ban forced Nvidia to write off US$5.5 billion in inventory and Mr Huang told the Stratechery podcast on May 19 that the company also had to walk away from US$15 billion in sales. The latest export restrictions introduced new limits on GPU memory bandwidth - a crucial metric measuring data transmission speeds between the main processor and memory chips. This capability is particularly important for AI workloads that require extensive data processing. Investment bank Jefferies estimates that the new regulations cap memory bandwidth at 1.7-1.8 terabytes per second. That compares with the 4 terabytes per second that the H20 is capable of. GF Securities forecast the new GPU will achieve approximately 1.7 terabytes per second using GDDR7 memory technology, just within the export control limits. REUTERS Join ST's Telegram channel and get the latest breaking news delivered to you.
Business Times
6 hours ago
- Business
- Business Times
Nvidia to launch cheaper Blackwell AI chip for China after US export curbs: sources
[BEIJING/TAIPEI] Nvidia will launch a new artificial intelligence (AI) chipset for China at a significantly lower price than its recently restricted H20 model and plans to start mass production as early as June, sources familiar with the matter said. The GPU or graphics processing unit will be part of Nvidia's latest generation Blackwell-architecture AI processors and is expected to be priced between US$6,500 and US$8,000, well below the US$10,000 to US$12,000 the H20 sold for, according to two of the sources. The lower price reflects its weaker specifications and simpler manufacturing requirements. It will be based on Nvidia's RTX Pro 6000D, a server-class graphics processor and will use conventional GDDR7 memory instead of more advanced high bandwidth memory, the two sources said. They added it would not use Taiwan Semiconductor Manufacturing Company's (TSMC) advanced Chip-on-Wafer-on-Substrate (CoWoS) packaging technology. The new chip's price, specifications and production timing have not previously been reported. BT in your inbox Start and end each day with the latest news stories and analyses delivered straight to your inbox. Sign Up Sign Up The three sources Reuters spoke to for this article declined to be identified as they were not authorised to speak to the media. An Nvidia spokesperson said the company was still evaluating its 'limited' options. 'Until we settle on a new product design and receive approval from the US government, we are effectively foreclosed from China's US$50 billion data centre market.' TSMC declined to comment. Market share plunge China remains a huge market for Nvidia, accounting for 13 per cent of its sales in the past financial year. It's the third time that Nvidia has had to tailor a GPU for the world's second-largest economy after restrictions from US authorities who are keen to stymie Chinese technological development. After the US effectively banned the H20 in April, Nvidia initially considered developing a downgraded version of the H20 for China, sources have said, but that plan did not work out. Nvidia CEO Jensen Huang said last week the company's older Hopper architecture – which the H20 uses – can no longer accommodate further modifications under current US export restrictions. Reuters was unable to determine the product's final name. Chinese brokerage GF Securities said in a note published on Tuesday (May 20) that the new GPU would likely be called the 6000D or the B40, though it did not disclose pricing or cite sources for the information. According to two of the sources, Nvidia is also developing another Blackwell-architecture chip for China that is set to begin production as early as September. Reuters was not immediately able to confirm specifications of that variant. Nvidia's market share in China has plummeted from 95 per cent before 2022, when US export curbs that impacted its products began, to 50 per cent currently, Huang told reporters in Taipei this week. Its main competitor is Huawei which produces the Ascend 910B chip. Huang also warned that if US export curbs continue, more Chinese customers will buy Huawei's chips. The H20 ban forced Nvidia to write off US$5.5 billion in inventory and Huang told the Stratechery podcast on Monday that the company also had to walk away from US$15 billion in sales. The latest export restrictions introduced new limits on GPU memory bandwidth – a crucial metric measuring data transmission speeds between the main processor and memory chips. This capability is particularly important for AI workloads that require extensive data processing. Investment bank Jefferies estimates that the new regulations cap memory bandwidth at 1.7 to 1.8 terabytes per second. That compares with the four terabytes per second that the H20 is capable of. GF Securities forecast the new GPU will achieve approximately 1.7 terabytes per second using GDDR7 memory technology, just within the export control limits. REUTERS
Time of India
20 hours ago
- Business
- Time of India
Nvidia to launch cheaper Blackwell AI chip for China after US export curbs
Nvidia will launch a new artificial intelligence chipset for China at a significantly lower price than its recently restricted H20 model and plans to start mass production as early as June, sources familiar with the matter said. The GPU or graphics processing unit will be part of Nvidia's latest generation Blackwell-architecture AI processors and is expected to be priced between $6,500 and $8,000, well below the $10,000-$12,000 the H20 sold for, according to two of the sources. The lower price reflects its weaker specifications and simpler manufacturing requirements. It will be based on Nvidia's RTX Pro 6000D, a server-class graphics processor and will use conventional GDDR7 memory instead of more advanced high bandwidth memory, the two sources said. They added it would not use Taiwan Semiconductor Manufacturing Co's advanced Chip-on-Wafer-on-Substrate (CoWoS) packaging technology. The new chip's price, specifications and production timing have not previously been reported. The three sources Reuters spoke to for this article declined to be identified as they were not authorised to speak to media. An Nvidia spokesperson said the company was still evaluating its "limited" options. "Until we settle on a new product design and receive approval from the U.S. government, we are effectively foreclosed from China's $50 billion data center market." TSMC declined to comment. Market share plunge China remains a huge market for Nvidia, accounting for 13% of its sales in the past financial year. It's the third time that Nvidia has had to tailor a GPU for the world's second-largest economy after restrictions from U.S. authorities who are keen to stymie Chinese technological development. After the U.S. effectively banned the H20 in April, Nvidia initially considered developing a downgraded version of the H20 for China, sources have said, but that plan didn't work out. Nvidia CEO Jensen Huang said last week the company's older Hopper architecture - which the H20 uses - can no longer accommodate further modifications under current U.S. export restrictions. Reuters was unable to determine the product's final name. Chinese brokerage GF Securities said in a note published on Tuesday that the new GPU would likely be called the 6000D or the B40, though it did not disclose pricing or cite sources for the information. According to two of the sources, Nvidia is also developing another Blackwell-architecture chip for China that is set to begin production as early as September. Reuters was not immediately able to confirm specifications of that variant. Nvidia's market share in China has plummeted from 95% before 2022, when U.S. export curbs that impacted its products began, to 50% currently, Huang told reporters in Taipei this week. Its main competitor is Huawei which produces the Ascend 910B chip. Huang also warned that if U.S. export curbs continue, more Chinese customers will buy Huawei's chips. The H20 ban forced Nvidia to write off $5.5 billion in inventory and Huang told the Stratechery podcast on Monday that the company also had to walk away from $15 billion in sales. The latest export restrictions introduced new limits on GPU memory bandwidth - a crucial metric measuring data transmission speeds between the main processor and memory chips. This capability is particularly important for AI workloads that require extensive data processing. Investment bank Jefferies estimates that the new regulations cap memory bandwidth at 1.7-1.8 terabytes per second. That compares with the 4 terabytes per second that the H20 is capable of. GF Securities forecast the new GPU will achieve approximately 1.7 terabytes per second using GDDR7 memory technology, just within the export control limits.
