Connect with us

Computers

Intel Splits Graphics Chip Unit Into Two, Consumer Graphics Division to Join Client Computing Group

Avatar

Published

on

Intel is doubling down on accelerated computing, a growing segment dominated by Nvidia as AI use surges.
By Reuters | Updated: 22 December 2022

Intel is splitting its graphic chips unit into two, the company said on Wednesday, as it realigns the business to better compete with Nvidia and Advanced Micro Devices.

The consumer graphics unit will be combined with Intel’s client computing group, which makes chips for personal computers, while accelerated computing teams will join its data centre and artificial intelligence (AI) business, the company said.

The move comes as Intel doubles down on accelerated computing, a growing segment dominated by Nvidia as AI use surges.

“I don’t think it changes much (if anything) other than aligning the products with the respective sales organizations they fit with vs. having them as a discrete segment,” Wedbush Securities analyst Matthew Bryson said.

Raja Koduri, who led the graphic chips unit, will return to his role as chief architect and oversee the company’s long-term technology and chip design strategy.

Koduri, who has led graphics technology ventures at iPhone maker Apple and AMD, joined Intel in 2017.

Earlier this month, it was reported that Intel had backed away from its original target of opening a chip factory in the eastern German city of Magdeburg in the first half of 2023. Regional newspaper Volksstimme reported that the semiconductor giant wanted more public subsidies.

The plant is central to German and European Union plans to strengthen the continent’s resilience by doing more manufacturing locally after the COVID-19 pandemic and Russia’s invasion of Ukraine highlighted the risks of long, globe-spanning supply chains.

But the newspaper said that surging energy and raw materials prices had upset the US company’s original calculations. Where Intel had originally budgeted for costs of EUR 17 billion (USD 18 billion, roughly Rs 1,48,000 crore), prices were now closer to EUR 20 billion (roughly Rs 1,76,000 crore), the paper said.

Computers

TSMC Defies Broader Industry Downturn, Posts 78 Percent Rise in Q4 Net Profit: All Details

Avatar

Published

on

TSMC's dominance in making some of the world's most advanced chips has kept the firm's order book full, according to analysts.
By Reuters | Updated: 12 January 2023

Taiwanese chipmaker TSMC posted a 78 percent rise in fourth-quarter net profit on Thursday, posting yet another quarterly record, as strong sales of advanced chips helped it defy a broader industry downturn that battered cheaper commodity chips.

Taiwan Semiconductor Manufacturing (TSMC), the world’s largest contract chipmaker and a major Apple supplier, saw net profit for October-December hit a record TWD 295.9 billion (roughly Rs.79,285 crore) from TWD 166.2 billion (roughly Rs. 44,540 crore) a year earlier.

That compared with the TWD 289.44 billion (roughly Rs. 77,570 crore) average of 21 analyst estimates compiled by Refinitiv.

TSMC’s business has been boosted by a global chip shortage sparked by pandemic-fuelled sales of smartphones and laptops. While the shortage has eased, analysts said dominance in making some of the world’s most advanced chips has kept the firm’s order book full.

Revenue for the quarter climbed 26.7 percent to $19.93 billion (roughly Rs. 5,340 crore), versus TSMC’s prior estimated range of $19.9 billion to $20.7 billion (roughly Rs. 5,545 crore).

TSMC’s share price fell 27.1 percent in 2022, but is up 8.5 percent so far this year giving the firm a market value of $412.78 billion (roughly Rs. 33,66,055 crore). The stock rose 0.4 percent on Thursday versus a 0.1 percent fall for the benchmark index.

Overall, the chip sector has been grappling with weak demand for gadgets such as smartphones as inflation accelerates and interest rates rise, against a backdrop of geopolitical tension.

In October, TSMC cut its annual investment budget by at least 10 percent for 2022 and struck a more cautious note than usual on upcoming demand, flagging challenges from rising inflationary costs and predicting a chip downturn for 2023.

The firm said it spent $36.29 billion (roughly Rs. 2,95,940 crore) on capital expenditure in 2022, compared to a previous forecast of around $36 billon.

TSMC, Asia’s most-valuable listed firm, whose clients include chip majors such as Qualcomm, has repeatedly said business would continue to be boosted by a “mega-trend” in the industry, brought by demand for high-performance computing chips for 5G networks and data centres, as well as increased use of chips in gadgets and vehicles.

© Thomson Reuters 2023

Continue Reading

Computers

Nvidia GeForce RTX 4070 Ti Specifications Accidentally Leaked Ahead of Launch: Report

Avatar

Published

on

The leaked specifications of Nvidia's RTX 4070 Ti GPU match those of the RTX 4080 with 12GB of memory, 7,680 CUDA cores, and a 2.61GHz boost clock.
By ANI | Updated: 2 January 2023

Nvidia has been tipped to release its “unlaunched” 12GB Nvidia RTX 4080 graphics card as the RTX 4070 Ti, and a new leak has given enthusiasts an idea of what to expect from the company’s graphics card. According to a report, the company briefly posted the specifications for its upcoming RTX 4070 Ti GPU on its website. The company removed the page, but a Twitter user was able to capture a screenshot before it was taken down.

According to a report by The Verge, before Nvidia pulled the page down, Twitter user @momomo_us was able to capture a screenshot, which was shared to the microblogging platform. So far, the leaked specifications match those of the 12GB RTX 4080, with 7,680 CUDA cores, a 2.61GHz boost clock, and 12GB of memory.


GeForce RTX 4070 Ti
Coming Month XXhttps://t.co/Z6Hv1Vv7ol pic.twitter.com/7sudSwgsFZ— 188号 (@momomo_us) December 30, 2022

Additionally, it claims that the GPU is capable of supporting 4K video at up to 240Hz and 8K at 60Hz with DSC and HDR, as per the report.

A chart included shows that the RTX 4070 Ti might outperform the RTX 3080 by almost 3.5 times when playing Cyberpunk 2077 in the game’s new Ray-Tracing: Overdrive mode.

In October, Nvidia faced criticism over its decision to launch the 12GB RTX 4080 GPU under the RTX 4080 moniker because of how much it differs from its much more powerful 16GB counterpart. It led Nvidia to cancel its launch altogether and plan a way to repackage the chip.

According to the report, pricing for the RTX 4070 Ti hasn’t yet been confirmed, but some rumours indicate that it will be cheaper than the $899 (roughly Rs. 74,200) 12GB RTX 4080. Nvidia is expected to launch the GPU at CES in January.

Continue Reading

Computers

China’s SMIC Ramps Up Production of Decade-Old 28nm Chips, US Lawmakers Raise Concerns

Avatar

Published

on

SMIC is doubling down on mature technology chips and has announced four new facilities as US export controls make it impossible to produce advanced chips.
By Reuters | Updated: 14 December 2022 11:01 IST

China’s largest chip maker SMIC is ramping up production of a decade-old chip technology, key to many industries’ supply chains, setting off alarm bells in the United States and prompting some lawmakers to try to stop them.

The United States and allied nations could further step up restrictions if China announces a $144 billion support package for its chip industry, as Reuters exclusively reported on Tuesday, said TechInsights’ chip economist Dan Hutcheson.

Starting with the Trump administration, the United States has been tightening the noose around China’s high-tech ambitions. It cut off the world’s largest telecommunications firm Huawei Technologies from the US market and technologies, as well as cut off air supply to China’s advanced chip making through a series of rules this year.

But why worry about older chip technology?

China, which in 2020 had 9 percent of the global chip market, has a track record of dominating key technologies by flooding the market with cheaper products and wiping out global competition, say China watchers.

They did it with solar panels and 5G telecom equipment, and could do it with older technology chips, said Matt Pottinger, former Deputy National Security Advisor of the United States during the Trump administration who has been studying chip policy at the Hoover Institution.

“It would give Beijing coercive leverage over every country and industry — military or civilian — that depend on 28-nanometer chips, and that’s a big, big chunk of the chip universe,” he said.

“28 nanometer” refers to a chip technology commercially used since 2011. It is still widely used in automotive, weapons and the explosive category of internet of things gadgets, said Hutcheson.

Hutcheson, who has been monitoring chip production capacity for four decades, said the concern is that Semiconductor Manufacturing International and other chipmakers in China could use government subsidies to sell chips at a low price. And a possible new round of financial support from Beijing would increase chip production even further.

“The Chinese could just flood the market with these technologies,” he said. “Normal companies can’t compete, because they can’t make money at those levels.”

US Lawmakers pushing against SMIC

Those concerns have pushed some lawmakers to use legislation for setting the defence budget hold back SMIC.

While the measure is weaker than what was initially proposed, this week US Senators are expected to pass the annual National Defense Authorization Act 2023 that includes a section barring the US government from using chips from SMIC and two other Chinese memory chip makers. It is not clear what impact the restriction, which kicks in five years after it becomes law, will have on SMIC.

Founded in 2000 with backing from Beijing, SMIC has long struggled to break into the ranks of the world’s leading chip manufacturers.

But it is a giant in older technology, including chips that regulate power flows in electronics. And its revenue was close to $2 billion in the third quarter this year, roughly double the same period last year on the back of the global chip shortage.

SMIC filing supply gasp

With U.S. export controls making it impossible to produce advanced chips, SMIC is doubling down on mature technology chips and has announced four new facilities, or fabs, since 2020. When those come online, it would more than triple the company’s output, estimates Samuel Wang, Gartner chip analyst. He said there is a huge ramp up in new chip fabs across China.

“All this will start to have an impact from early 2024 and will be full blown by 2027,” said Wang, adding the chip supply increase will put downward pressure on chip prices.

The importance of older chip technology hit the industry in the face in 2021 as a shortage of those chips prevented manufacturing of millions of cars and consumer electronics.

Mark Li, Bernstein Research’s chip analyst in Asia, said the company is becoming a formidable competitor to Taiwan’s UMC Microelectronics and US-headquartered GlobalFoundries.

“SMIC has been much more willing to add capacity than other fabs at the low-end, and especially in this shortage we’ve seen in the past two years,” he says. “It’s not an issue now…but who knows, maybe in a few years there will be another shortage and capacity will be a big problem.”

© Thomson Reuters 2022

Continue Reading

Computers

TSMC Expects Annual Revenue of $10 Billion From US Chip Plants; Apple to Use New Chips, Tim Cook Says

Avatar

Published

on

TSMC says it is more than tripling its planned investment in the chip factories to $40 billion (roughly Rs. 330 crore)
By Reuters | Updated: 7 December 2022

TSMC on Tuesday estimated annual revenue of $10 billion when its two planned chips fabrication plants open in Arizona. The Taiwanese chipmaker said it was more than tripling its planned investment in the factories to $40 billion. The first fab will be operational by 2024 while the second facility nearby will produce advanced chips by 2026. US President Joe Biden and others, including the CEOs of major TSMC customers, are attending a “tool-in” ceremony for the symbolic moving of the first equipment onto the shop floor of the new $12 billion (roughly Rs. 99 crore) facility.

“When completed with both fabs, we will manufacture over 600,000 wafers a year, representing $10 billion (roughly Rs. 82 crore) yearly revenue and with our customers product sales over $40 billion (roughly Rs. 330 crore) a year,” said TSMC Chief Executive Mark Liu.

The projects will result in 31,000 construction jobs and “create an additional 13,000 high pay high tech jobs including the 4,500 direct TSMC employees,” Liu added.

Apple, Nvidia, and Advanced Micro Devices, all major TSMC customers, said they expected their chips to be made in the new Arizona plants.

“We work with TSMC to manufacture the chips that help power our products all over the world. And we look forward to expanding this work in the years to come as TSMC forms new and deeper roots in America,” said Apple CEO Tim Cook.

“AMD expects to be a big customer, of both fabs and we’re committed to working closely with TSMC and the entire ecosystem,” said AMD CEO Lisa Su.

At least a dozen major cranes are still set up around the first factory which is dubbed Fab 21. The factory is in the northern part of Phoenix, surrounded by brown hills and empty land.

With the new TSMC factory in the backdrop with the flag and a drape reading “A Future Made in America Phoenix, AZ,” TSMC executives led by founder Morris Chang, 91, along with CEOs of key machine suppliers and Apple, Nvidia and AMD, toasted the factory opening with sparkling wine.

At least a dozen major cranes are still set up around the first factory which is dubbed Fab 21. The factory is in the northern part of Phoenix, surrounded by brown hills and empty land.

With the new TSMC factory in the backdrop with the flag and a drape reading “A Future Made in America Phoenix, AZ,” TSMC executives led by founder Morris Chang, 91, along with CEOs of key machine suppliers and Apple, Nvidia and AMD, toasted the factory opening with sparkling wine.

© Thomson Reuters 2022

Continue Reading

Computers

US Senators Lobbying for Ban on Government Deals With Chinese Chipmakers: Report

Avatar

Published

on

By Reuters | Updated: 18 November 2022

US Senate Majority Leader Chuck Schumer and Republican Senator John Cornyn are lobbying hard for a ban on government business with Chinese chipmakers, Politico reported on Thursday, citing three people familiar with the matter. The senators want to get their amendment which blocks federal access to semiconductor products and services made by Chinese firms into the final version of this year’s National Defense Authorization Act (NDAA), the report said.

The measure would broaden provisions in Section 889 that already prohibit government agencies from doing business with Chinese telecommunications companies or contractors who use their technologies, according to a report by Politico.

Schumer and Cornyn got their proposal added to the Senate NDAA last month in the October managers package and are now working to convince their colleagues, the report added.

The fiscal 2023 NDAA must pass the Senate and House of Representatives later this year before it can be sent to the White House for US President Joe Biden to sign into law.

Last month, the Biden administration published a sweeping set of export controls, including a measure to cut off China from certain semiconductor chips made anywhere in the world with US tools.

On August 9, Biden signed a landmark bill to provide $52.7 billion (roughly Rs. 430 crore) in subsidies for US semiconductor production and research and to boost efforts to make the US more competitive with China’s science and technology efforts.

At the time, Biden had touted investments that chip companies are making even though it remains unclear when the US Commerce Department will write rules for reviewing grant awards and how long it will take to underwrite projects.

The legislation authorises $200 billion (roughly Rs. 16,34,700 crore) over 10 years to boost US scientific research to better compete with China. Congress would still need to pass separate appropriations legislation to fund those investments.

© Thomson Reuters 2022

Continue Reading

Computers

Nvidia Working With Microsoft to Build ‘Massive’ Cloud AI Computer Using Several Thousand GPUs

Avatar

Published

on

By Reuters | Updated: 17 November 2022

US chip designer and computing firm Nvidia on Wednesday said it is teaming up with Microsoft to build a “massive” computer to handle intense artificial intelligence computing work in the cloud.

The AI computer will operate on Microsoft’s Azure cloud, using tens of thousands of graphics processing units (GPUs), Nvidia’s most powerful H100 and its A100 chips. Nvidia declined to say how much the deal is worth, but industry sources said each A100 chip is priced at about $10,000 (roughly Rs. 8,15,400) to $12,000 (roughly Rs. 9,78,500), and the H100 is far more expensive than that.

“We’re at that inflection point where AI is coming to the enterprise and getting those services out there that customers can use to deploy AI for business use cases is becoming real,” Ian Buck, Nvidia’s general manager for Hyperscale and HPC told Reuters. “We’re seeing a broad groundswell of AI adoption … and the need for applying AI for enterprise use cases.”

In addition to selling Microsoft the chips, Nvidia said it will partner with the software and cloud giant to develop AI models. Buck said Nvidia would also be a customer of Microsoft’s AI cloud computer and develop AI applications on it to offer services to customers.

The rapid growth of AI models such as those used for natural language processing have sharply boosted demand for faster, more powerful computing infrastructure.

Nvidia said Azure would be the first public cloud to use its Quantum-2 InfiniBand networking technology which has a speed of 400 gigabits per second. That networking technology links servers at high speed. This is important as heavy AI computing work requires thousands of chips to work together across several servers.

© Thomson Reuters 2022

Continue Reading

Trending