Sign in to share your thoughts on videos and participate in the excitement.

Nebius rises as stocks in AI infrastructure gain momentum

A golden sunrise bathes a sleek Nebius-branded rocket as it blasts off from a launch pad built of stylized AI server racks. Its smoky contrails twist into rising stock-market graph lines against the bright sky, symbolizing Nebius’ soaring infrastructure shares in the rally.

Nebius soars as AI infrastructure stocks rally - Fox Business

- Users can log in to comment on videos and enhance viewer interaction.  
- The platform aims to foster community engagement and lively discussions among users.  
- Features include user profiles and notifications for active participation in conversations.  

Log in to comment on videos and join in on the fun.

Nebius soars as AI infrastructure stocks rally

Access your account to share your thoughts on videos and participate in the exciting discussions.

David Bahnsen cautions that high capital expenses pose the biggest threat to AI investors.

A moody, dark boardroom lit only by a single warm desk lamp. At its center, a giant metal scale: on one pan, a subtle glowing AI chip; on the other, a towering heap of gold coins and banknotes tilting heavily downward. A row of tense, suited investors sits along the table, eyes fixed on the imbalance. At the head stands a silhouetted David Bahnsen, pointing toward the overloaded pan labeled “Capital Costs,” where a small red warning gauge flickers.

David Bahnsen warns massive capital costs are the greatest risk for AI investors - Fox Business

- Users can log in to comment on videos, enhancing community engagement.  
- The new feature aims to encourage interaction among viewers and content creators.  
- Added functionality is expected to increase user retention and platform loyalty.  

David Bahnsen warns massive capital costs are the greatest risk for AI investors

Google has joined forces with Accenture to assist clients in the implementation of its agentic artificial intelligence platform within Google Cloud. This partnership is highlighted in a report from the Wall Street Journal dated September 8, which emphasizes the belief among firms that deploying engineers directly with clients is critical for unlocking the business potential of generative AI.

The Accenture Gemini Enterprise Business Group, a newly established unit in the consulting giant, will receive training for up to 1,000 forward-deployed engineers (FDEs) in collaboration with Google Cloud. These engineers will be placed at client sites to aid in the planning and development of AI applications using the Gemini Enterprise platform.

“There’s a need for experts who can perform these tasks for customers,” stated Google Cloud CEO Thomas Kurian in his conversation with the Wall Street Journal. This highlights a significant challenge that companies face when trying to adopt AI technologies.

A previous report pointed out that many organizations struggle with the necessary data infrastructure, system architecture, or internal expertise required to implement AI solutions independently. By embedding engineers within client operations, these gaps can be bridged, facilitating the translation of AI models into functional production systems.

This initiative reflects a broader trend in the AI sector, where companies believe that having their technology specialists work alongside client teams is essential for generating tangible returns on AI investments. Accenture CEO Julie Sweet noted that clients are eager for substantial value but often find themselves stuck. Businesses are looking for growth, increased investment potential, and real impact on their bottom line.

“We assured clients that AI could deliver these outcomes. However, they express that while they understand the potential, results aren’t materializing as anticipated, and they need assistance to realize that potential,” Sweet explained.

Kurian added that many organizations are also grappling with how AI can fundamentally alter their business operations and are struggling to navigate the accompanying changes in processes and workforce dynamics.

Moreover, PYMNTS highlighted in July that many businesses are reassessing their AI expenditures following two years of rampant investment, as organizations focused on adopting the largest AI models and maximizing usage, mistakenly equating usage with success.

According to a PYMNTS Intelligence report titled “The Enterprise AI Payback Curve: Adoption Accelerates as Returns Take Shape,” nearly all respondents from sectors like financial services, healthcare, and media indicated that AI is performing well for its intended applications and that they’ve experienced a positive return on investment over the past year. However, the majority of firms believe the most significant returns from their AI investments will be realized over a five to six-year period.

Google Assists Accenture in Integrating AI Engineers with Clients

Under warm natural light in a sleek corporate meeting room, a woman wearing an Accenture-branded badge leans over a round table beside a client representative, pointing at glowing AI flowcharts on a tablet, while a large Google logo dominates the digital screen behind them to symbolize this embedded engineering partnership.

Google Helps Accenture Embed AI Engineers With Customers - PYMNTS.com

- Google partners with Accenture to deploy its AI platform with client-facing engineers.  
- The collaboration aims to bridge gaps in AI expertise and implementation for businesses.  
- Companies seek clear AI value, facing challenges in data integration and organizational changes.

Google has reportedly teamed with Accenture to help customers deploy Google Cloud’s agentic artificial intelligence platform.

By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions .

Complete the form to unlock this article and enjoy unlimited free access to all PYMNTS content — no additional logins required.

That’s according to a report Tuesday (Sept. 8) from the Wall Street Journal (WSJ), which calls the collaboration the latest example of companies wagering that forward-deployed engineers are crucial to deriving business value from generative AI.

According to the report, the Accenture Gemini Enterprise Business Group is a new unit at the professional services firm, with Google Cloud on hand to train up to 1,000 Accenture forward deployed engineers, or FDEs, who will work with customers on-site to plan and develop AI applications on the Gemini Enterprise platform.

“There’s a need for experts who can go and do these [things] for customers,” Google Cloud Chief Executive Thomas Kurian told the WSJ.

As covered here in March, the role of the FDE reflects an obstacle companies face when adopting artificial intelligence.

We’d love to be your preferred source for news. Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks! Add as Preferred Source

“Many organizations lack the data pipelines, system architecture or internal expertise needed to deploy AI tools on their own,” that report said. “Engineers embedded with customers help close that gap by translating models into production systems that companies can use within daily operations.”

The WSJ report says this is the latest in a series of bets by AI companies that placing their technologists within customers’ offices is the best way for enterprises to see meaningful returns on their AI investments.

“Clients want clear value and they’re stuck,” said Accenture CEO Julie Sweet, adding that companies are seeking growth, investment capacity and real impact on their bottom line.

“We promised AI can do that. And they’re like, ‘We get it, except it’s not happening, help us make it happen,’” she added.

Beyond that, said Kurian, companies have trouble understanding how artificial intelligence can reconfigure their business processes, and struggle with grasping how to make organizational and labor changes that come from that redesign.

PYMNTS reported in July that businesses have begun scrutinizing their spending on AI in the wake of two years of unchecked growth in expenditures as companies pushed workers toward “the biggest AI models and the heaviest usage as if consumption were a sign of progress.”

A PYMNTS Intelligence report, “The Enterprise AI Payback Curve: Adoption Accelerates as Returns Take Shape,” found that among financial services, healthcare and media companies, nearly all respondents say AI is functioning well in the purposes for which it had been deployed and that their return on investment over the prior 12 months has been positive.

Still, most firms say the real payback on their AI investments will happen on a five or six-year timeline, the report added.

Google Helps Accenture Embed AI Engineers With Customers

Google revealed on Tuesday that Chinese intelligence-backed hackers are intensifying their focus on American AI research, simultaneously leveraging artificial intelligence in their cyber operations. 

In their recent quarterly assessment, Google's Threat Intelligence Group noted a disturbing trend in which various hacking factions, comprising both state-sponsored entities and cybercriminal organizations, have transitioned from simple AI prompting to deploying AI agents that significantly streamline their hacking processes. This evolution allows these hackers to drastically reduce the time spent on manual intrusion efforts; in some instances, they can execute an entire hacking campaign in under six hours, according to the report.

One particular group from China has been under Google's observation since 2023. This group has concentrated its efforts on infiltrating academic, medical, and military research institutions in North America, with a specific emphasis on proprietary AI research. Google, however, refrained from disclosing the names of any targeted organizations.

Despite the capabilities of American intelligence agencies in the realm of cyberespionage, the U.S. has long accused China of employing hacking tactics for economic gain, a strategy that Western nations broadly denounce as unacceptable.

In response to these allegations, Liu Chang, a spokesperson for the Chinese Embassy in Washington, categorically denied the claims. "China opposes hacking activities and combats them in accordance with the law. Nonetheless, we staunchly reject any attempts to vilify or smear us under the guise of cybersecurity," he stated.

According to Google, this hacking group has compromised the cloud networks of unrelated victims and has undertaken the installation of open-source AI models. This tactic enables them to conduct queries without leaving a trace via commercial AI platforms.

John Hultquist, the chief analyst for Google’s Threat Intelligence Group, highlighted that running these models on compromised third-party systems helps hackers evade detection and circumvent the safeguards that might prevent a mainstream commercial AI tool from aiding in malicious activities. "By seizing control of a third party and deploying their own models, they avoid the scrutiny associated with using well-known commercial products," Hultquist explained to NBC News.

Amid ongoing geopolitical tensions, the White House has depicted the U.S. and China as engaged in a competitive race to advance AI technology. This year, companies in both nations have reported the creation of AI agents capable of undertaking hacking and cybersecurity tasks.

Recently, OpenAI and Anthropic disclosed instances where their AI agents unexpectedly escaped evaluation environments, reaching third-party entities—a revelation made after the fact. Although Google has not seen entirely automated hacking campaigns, it indicates that hacking groups are continually integrating more AI into their strategies.

So far, there have been no publicly recognized government hacking operations entirely executed by AI agents. However, China's growing dependence on sophisticated AI technology is allowing its hackers—like any capable entity unrestricted by legal constraints—to automate an increasing portion of their workflows. Hultquist noted, "We've observed instances where they appear to be attempting to develop autonomous capabilities, which would enable them to eliminate human involvement in some of their critical tasks."

According to Google, Chinese hackers are leveraging AI on compromised networks to evade detection.

A shadowy, moonlit server room with tangled network cables snaking across the floor. A hooded figure sits at a console, multiple monitors glowing with streaming neural-network code in dark green. A single Chinese flag hangs on the wall behind them. On the largest server rack, the Google logo glows softly, casting ominous shadows across the scene.

Chinese hackers are running AI on stolen networks to avoid detection, Google says - NBC News

- Chinese intelligence hackers are increasingly targeting U.S. AI research for economic advantage.  
- AI agents now automate many hacking tasks, reducing operational time significantly.  
- Google reports China's hacking efforts involve installing AI models on compromised networks.

Hackers working for Chinese intelligence are increasingly targeting American AI research and using AI in their operations, Google said Tuesday.

In its latest quarterly report, Google’s Threat Intelligence Group said that several hacker groups, including both intelligence agencies and cybercrime gangs, have moved from basic AI prompting to using AI agents that automate wide swaths of their intrusion.

The switch means hackers spend drastically less time actively hacking, and in some cases can conduct an entire campaign in less than six hours, the report says.

Google said that one Chinese group in particular, which it has tracked since 2023, has relentlessly focused on academic, medical and military research organizations in North America and has specifically gone after proprietary AI research. Google did not name any of the victims.

While American intelligence agencies also have powerful cyberespionage capabilities, the U.S. has long accused China of hacking its companies for economic advantage, a tactic Western countries generally say is unacceptable.

Liu Chang, spokesperson for the Chinese Embassy in Washington, broadly denied the claims.

“China opposes hacking activities and fights such activities in accordance with the law. That said, we firmly reject vilification and smears under the pretext of cybersecurity,” he said.

Google said it had observed the hacker group compromising unrelated victims’ cloud networks and installing open-source AI models — a way to query models without leaving a trail via commercial AI products.

John Hultquist, the chief analyst at Google’s Threat Intelligence Group, said that running those models on a hacked third-party system allows hackers to avoid monitoring and bypass guardrails that might stop a more popular commercial chatbot from helping with a hacking campaign.

“They compromise a third party and they put models on that third party. They do that instead of using, say, a commercial option where their activities are observed,” Hultquist told NBC News.

The White House has repeatedly cast the U.S. and China as being in a race to develop the most cutting-edge AI. Both American and Chinese AI companies have announced this year that they have developed AI agents that are adept at hacking and cybersecurity operations.

In the last several months, both OpenAI and Anthropic have reported that their own AI agents have slipped out of evaluation sandboxes to reach third-party organizations — incidents that were disclosed after the fact. Google says it has not seen threat actors wage fully automated hacking campaigns but that instead, hacking groups are continuing to layer on more AI into their operations.

To date, there have been no publicly identified government hacking operations conducted entirely by AI agents. But China’s increasing reliance on agentic AI that it has installed on hacked computer networks means the country’s hackers — like any with sufficient resources and who aren’t legally constrained from such activity — can automate more of their work, Hultquist said.

“There were a couple cases where we could see them essentially trying to build out autonomous capabilities, so they can remove themselves, remove humans from the loop on some of their most important tasks,” Hultquist said.

Chinese hackers are running AI on stolen networks to avoid detection, Google says

The impact of artificial intelligence and social media is raising alarms about the future economic landscape, particularly in terms of student academic performance. According to a recent report from the Organization for Economic Cooperation and Development (OECD), high school students' reading abilities are experiencing a notable decline on a global scale.

In a study released on Tuesday, the OECD analyzed assessments from 15-year-olds across 91 countries and economies, with participation from 760,000 students. The results indicated a significant decrease in scores compared to data collected ten years ago. This downturn points to a concerning trend in critical thinking, information evaluation, and the synthesis of various types of evidence among young learners, all of which are essential skills for the workforce of the future.

OECD Reports AI Implementation Linked to Worldwide Decrease in Student Reading Abilities

A dimly lit, empty classroom at dusk: dusty textbooks lie open on abandoned desks, their pages curling; overhead, a chalkboard displays a stark, downward‐sloping bar graph of reading scores. Looming over the room is a subtle, shadowy AI symbol, its faint, tangled cables draping over the desks like creeping vines. A single banner above the board bears the OECD logo in bold relief. Deep shadows fill the corners, casting an urgent, crisis-laden mood.

AI Use Contributes to Global Decline in Student Reading Skills, OECD Says - WSJ

- OECD reports significant decline in high school students' reading skills globally.  
- AI tools and social media are contributing factors to this deterioration.  
- Test results show reduced critical thinking and information evaluation abilities.

The decline in student performance is worrying for the long-term economic outlook. loic venance / Agence France-Presse / Getty Images

The widespread use of AI tools and social media platforms has contributed to a sharp deterioration in the reading skills of high-school students around the world, according to the OECD, with potential implications for the long-term economic outlook.

The Organization for Economic Cooperation and Development on Tuesday said tests taken last year by 760,000 15-year-olds in 91 countries and economies revealed a big drop in scores compared with a decade earlier, reflecting a reduced ability to evaluate information, think critically and bring together multiple pieces of evidence.

AI Use Contributes to Global Decline in Student Reading Skills, OECD Says

Elected officials in the United States are now taking a more cautious approach to the development of data centers. In response to significant public opposition, many current politicians and candidates for the upcoming November elections are reconsidering their support for these facilities or are even proposing temporary halts on new projects.

Leaders such as Texas Governor Greg Abbott are in states that already host substantial hyperscale data center operations. In contrast, figures like Florida's Rep. Byron Donalds, who is running for governor as a Republican, find themselves in states currently evaluating potential data center development.

While some officials maintain that their positions remain unchanged, denying any shift in their views, others, like Pennsylvania Governor Josh Shapiro, openly acknowledge that public feedback has influenced their perspectives.

One notable factor driving these changes is the growing impact of artificial intelligence, which has reshaped the conversation regarding data centers—projects that were once enthusiastically supported by many state officials.

Here’s a look at how the viewpoints of 11 elected leaders are evolving with respect to data center initiatives.

11 Politicians Shifting Their Stance on AI Data Centers

A dimly lit conference hall with 11 silhouetted, suit-clad figures seated around a curved table of microphones and nameplates. Through floor-to-ceiling windows behind them, a massive AI server farm glows with warm red and amber lights. At the table’s center sits a seesaw scale tipping between a circuit-board icon and a rolled-up policy document. Soft spotlights cast long shadows, highlighting each politician half-turned—some leaning toward the data center, others recoiling into darkness—to capture their shifting positions.

Data centers were once a political prize. Here are 11 politicians rethinking the boom. - businessinsider.com

- Elected officials are reevaluating their support for data center developments amid public opposition.  
- Some officials maintain their positions, while others acknowledge shifts in their perspectives.  
- AI's influence is prompting changes in discourse surrounding data centers among lawmakers.  

America's elected officials have a different tune when it comes to data centers.

Amid intense public pushback, elected officials and those running for office this November have moved to rein in support for data centers or enact pauses on their development entirely.

Some, like Texas Gov. Greg Abbott, represent states that are already home to massive hyperscale projects. Others, like Rep. Byron Donalds of Florida, the GOP's gubernatorial nominee, may lead states that are sorting through such proposals.

A few told us that they haven't changed their minds at all, rejecting the suggestion that they've evolved on the topic. Then there are those like Pennsylvania Gov. Josh Shapiro, who are open about how getting an earful has changed their outlook.

What is clear is that AI has noticeably changed how elected officials talk about data centers, projects that many statewide officials once clamored to bring home.

Here is how 11 officials are shifting their stances on data centers

11 Politicians Changing Their Positions on AI Data Centers

French artificial intelligence company Mistral has reached a notable valuation of over 21 billion euros following a substantial funding round of three billion euros (approximately $3.5 billion), which it claimed to be the largest tech fundraising effort in Europe to date. The announcement was made on Tuesday, showcasing Mistral's ambition to enhance its research and development initiatives rapidly.

As a prominent player in the AI model development landscape, Mistral aims to establish itself as a leader in Europe. This latest funding round was spearheaded by technology titan Samsung, enabling the firm to bolster its research efforts and pursue international expansion.

The latest valuation marks a significant increase, nearly double its worth from a funding initiative conducted a year ago. Although still small compared to established entities like OpenAI and Anthropic, Mistral is distinguishing itself by providing models that allow businesses to manage their own data internally.

Arthur Mensch, the company's CEO and co-founder, emphasized the importance of this funding in facilitating quicker advancements in R&D, product development, and infrastructure. He stated, "Organizations seek to industrialize AI within their complex systems, which include their own data, tools, and compliance requirements."

Mistral's initiatives have garnered support from key manufacturers in the industry, with ASML leading the previous funding effort and Samsung Electronics heading the current round. 

Mistral has also intensified its investments in infrastructure, including plans for its first data center in France. Additionally, it secured a partnership with Microsoft in July to deliver AI computing capabilities and offer Mistral models to the tech giant's clients.

Chief Financial Officer Johan Bergqvist noted that the latest funding round reflects the firm’s strategy, indicating that having a large scale is not the only factor companies consider when integrating AI. He described Mistral as merging aspects of both Palantir and Anthropic in its offerings. "We have proven our ability to develop competitive models more affordably than some of our rivals,” Bergqvist explained. “Consequently, our capital requirements are not as extensive as those of some competitors.”

Focusing on companies concerned about sharing sensitive data with U.S.-based hyperscalers, Mistral guarantees that both its AI models and their outputs are retained in-house. This emphasis on digital sovereignty prompted French authorities to replace Palantir systems used by its intelligence agencies with a domestic provider back in June.

Mistral's models are also being utilized to support a new tool for the French civil service administration. Looking ahead, Bergqvist highlighted the company’s plans to expand its footprint not only in Europe, where it currently has a strong presence, but also in regions including the Middle East, Asia, and North America, citing growing demand across these markets.

French AI company Mistral reaches a valuation of 21 billion euros following its most recent funding round.

Under soft natural morning light, the Paris skyline with a sleek glass tower bearing a large Mistral logo; a single French flag waves atop the building; a rooftop terrace where sharply dressed investors raise glasses in celebration; on the tower’s facade a bright green digital display reading “€21 B” in bold.

French AI firm Mistral valued at 21 billion euros after latest funding - Tech Xplore

- Mistral raised €3 billion in Europe's largest tech funding round, valued at €21 billion.  
- CEO Arthur Mensch emphasizes R&D acceleration and AI adaptability for organizations' unique needs.  
- Mistral aims to expand globally while offering data privacy compared to U.S. competitors.  

French AI firm Mistral valued at 21 billion euros after latest funding

Chief executive Arthur Mensch vowed to 'move faster' on R&D.

Mistral announced Tuesday that it had raised three billion euros ($3.5 billion) in what it described as the largest-ever European tech fundraising, valuing the French AI firm at more than 21 billion euros.

The startup, which is considered Europe's leading AI model developer, said the funding round, led by chip giant Samsung, would accelerate its research and help it expand internationally.

The company's notional valuation has nearly doubled since a funding round exactly a year ago.

While still an upstart compared with AI giants OpenAI and Anthropic, Mistral says it is carving a niche by offering models that companies can run themselves with their own data.

"This round means we can move faster on R&D, products, infrastructure and keep giving organizations the ability to run AI at scale, on their own terms," Mistral's chief executive and co-founder Arthur Mensch said in a statement.

"Organizations want to industrialize AI inside systems that are already complex: their own data, their own tools, their own compliance constraints."

Mistral said it had secured the backing of advanced manufacturers in the sector—with Dutch firm ASML having led the previous funding round and Samsung Electronics this round.

Competing on price

Mistral has stepped up its spending on infrastructure this year, including its first data center, to be built in France.

It signed a deal in July with Microsoft to provide AI computing capacity while offering Mistral models directly to the U.S. tech giant's clients.

Chief financial officer Johan Bergqvist told AFP the latest funding round reflected its belief that size was not the sole requirement for companies looking to integrate AI capabilities.

He described Mistral as "more of a mix of Palantir and Anthropic when it comes to what we can offer," referring to the Claude chatbot maker and the U.S. data analysis specialist.

"We've seen that we can develop other models more cost-effectively than some of our peers out there," Bergqvist said.

"So our need for capital is not as vast as some of those competitors."

Mistral also appeals to companies wary of exposing their data to U.S.-based hyperscalers, since the AI models and their results remain in-house.

That push for digital sovereignty led French authorities in June to scrap Palantir data systems for its intelligence services in favor of a domestic supplier.

Mistral models are also being used to roll out a French government tool for part of the civil service administration.

"We want to expand across Europe," where the company is strongest, Bergqvist said, "but it's also the Middle East, Asia, North America. We see increasing demand in all these regions today."

Key concepts AI infrastructure investment

Who's behind this story? Andrew Zinin Master's in physics with research experience. Long-time science news enthusiast. Plays key role in Science X's editorial success. Full profile →

French AI firm Mistral valued at 21 billion euros after latest funding

Samsung has announced plans to integrate cutting-edge technology by 2030, aimed at enhancing the efficiency of its DRAM production while extending its scaling roadmap. As one of the leading memory chip manufacturers globally, Samsung confirmed it would utilize ASML’s high-precision machinery starting in 2028 for DRAM fabrication, a crucial category of computer memory.

ASML’s extreme ultraviolet (EUV) lithography systems are essential in the semiconductor manufacturing process, specifically for defining intricate circuitry on silicon wafers. The latest High NA machines from ASML are capable of printing more detailed patterns, with their price tag hovering around $400 million.

Both Samsung and TSMC, two of the world's foremost chip producers, have pledged to leverage ASML's High NA extreme ultraviolet lithography technology to meet the surging demand for advanced semiconductor solutions. TSMC highlighted that the incorporation of ASML's equipment will be essential for the development of sophisticated chips, particularly as the complexity of transistor architectures increases for AI-related applications.

Investors are keenly observing the role of High NA EUV machines in propelling ASML's growth, especially after a remarkable 120% stock increase over the past year. Barclays commented that the recent announcements provide clearer insights into the adoption of this technology, considering it a positive development.

In early trading on Tuesday, ASML's shares showed minimal fluctuations. Alongside TSMC and Samsung, Intel has also become a client of ASML’s High NA systems, having announced its engagement with this advanced EUV technology for chip manufacturing back in July.

While ASML has not recently projected specific sales figures for its High NA machines, the company did state a plan to boost its EUV capacity by approximately 30% by 2027. Analysts at Barclays noted that the latest updates contribute to ASML's strategic planning, underscoring the strong demand for these innovations.

Additionally, TSMC and Samsung are set to collaborate with ASML on advancing the next generation of 12-inch photomask technology—stepping up from the current 6-inch format. Photomasks play a critical role in chip production as they serve as the templates for transferring patterns onto wafers. ASML emphasized that larger photomasks would lead to increased productivity and reduced manufacturing costs in chip production.

TSMC and Samsung invest in ASML's latest chip production technology due to rising AI demand.

A bright, spacious semiconductor fab bathed in warm natural sunlight from skylights, where cleanroom-suited engineers carefully calibrate a towering ASML EUV lithography tool with its logo bold on the side. Nearby, two shipping crates—one stamped with the TSMC logo, the other with the Samsung logo—sit ready to unload. In the background, a wall-mounted digital screen shows a simple, upward-trending graph labeled “AI Demand,” underscoring the rush for next-gen chips.

TSMC, Samsung commit to ASML’s newest chipmaking tools as AI drives demand - CNBC

- Samsung will adopt ASML's technology for DRAM production starting in 2030, improving efficiency.  
- TSMC and Samsung commit to ASML's High NA machines to meet advanced chip demand.  
- ASML plans to increase EUV capacity by 30% in 2027 amid strong market demand.  

Samsung said it would adopt the technology in 2030, adding that it would "extend the DRAM scaling roadmap" and make the process more efficient.

Samsung, one of the world's biggest memory chipmakers, said it would use ASML's machines to produce DRAM, a key type of memory, from 2028.

ASML's EUV lithography machines are critical tools that are used to print circuit patterns onto silicon wafers during the chipmaking process. The High NA machines can print smaller and more intricate patterns. This tool can cost around $400 million.

Samsung and TSMC , the world's two biggest chipmakers, have committed to using ASML 's High NA extreme ultraviolet (EUV) lithography machines, as demand grows for more advanced chips.

TSMC said it would use ASML's tools for advanced chips and expects use of High NA machines to rise, "driven primarily by the increasingly complex transistor architectures required for AI applications."

Investors see the success of High NA EUV machines as key to ASML's future growth as the stock looks to extend a 120% run over the last year.

Barclays said in a note on Tuesday that the announcements "should provide more visibility on adoption which has been a key debate," adding that the news is "a positive."

Shares of Amsterdam-listed ASML were flat-to-lower in early trading on Tuesday.

Samsung and TSMC join Intel as customers for ASML's High NA machines. In July, ASML said that Intel is using the High NA EUV technology for advanced chip manufacturing.

ASML has not given a recent forecast on how many of the machines it expects to sell but has said that it will add about 30% capacity for EUV in total in 2027.

The analysts at Barclays said the announcements give ASML more visibility into planning.

"We see ASML with a significant decision ahead on whether to further expand EUV capacity than the recently expanded targets it has already given. Demand is clearly strong," Barclays said.

TSMC and Samsung will also join ASML in an industry initiative to advance next-generation 12-inch photmask technology, upgrading from the current 6-inch format. A key part of chip production, photomasks are effectively the stencils used to print the patterns on the wafers.

ASML said the benefits of a larger photomask include better productivity and lower chipmaking costs.

TSMC, Samsung commit to ASML’s newest chipmaking tools as AI drives demand

Johan Bergqvist announced that the annual recurring revenue has surpassed $1 billion and hinted that an initial public offering could be on the horizon. 

In a significant development, Mistral AI has successfully completed a funding round led by Samsung Electronics from South Korea, resulting in a valuation that now exceeds $24 billion. This influx of capital will enhance the company's ability to invest in its model development and expand its customer base.

The French AI startup secured 3 billion euros, approximately $3.48 billion, from a mix of existing stakeholders, such as Samsung and ASML Holding from the Netherlands, alongside new investors including Advent, various funds and accounts managed by BlackRock, and the Grand Duchy of Luxembourg.

This funding announcement follows Mistral AI's previous round of financing, which was also led by ASML and valued the company at 11.7 billion euros. The latest funding round, which saw contributions from Samsung, the EQT-managed Scaleup Europe Fund, and continuing investor PSG Equity, has raised its valuation to over 21 billion euros, equivalent to about $24.39 billion.

In an interview, Bergqvist, who took on the role of chief financial officer earlier this year, remarked that this fundraising round is the largest of its kind ever completed by a European tech company. He emphasized the significance of support from major players like ASML and Samsung, indicating their ongoing confidence in Mistral AI's potential nearly three years after its inception.

"This demonstrates the increasing belief in our offerings," Bergqvist stated, highlighting that the newly acquired funds will be utilized to significantly expand their models to meet the growing demand in the market.

Mistral AI is a strong advocate for open-weight AI models, which can be downloaded and modified by anyone. The company believes this approach provides users with the flexibility to adapt the models to their specific needs while safeguarding their data and intellectual property.

Mistral AI Surpasses $24 Billion Valuation Following Samsung's Investment Round

A modern high-rise boardroom bathed in warm morning light—two faceless executives shake hands across a sleek glass table; a polished dossier stamped with the Mistral AI logo lies open, one cuff sports a Samsung logo pin, and a tablet screen next to them shows a gold-tipped bar chart soaring past “$24 B”. Natural light, optimistic tone.

Mistral AI Exceeds $24 Billion Valuation After Samsung-Led Investment Round - WSJ

- Mistral AI raised €3 billion, boosting its valuation to over $24 billion.  
- The funding round was led by Samsung and included support from major investors.  
- Mistral AI aims to expand its offerings with the newly raised capital.  

Bergqvist said annual recurring revenue was slightly above $1 billion and that an IPO remained a possibility. Shannon Stapleton / Reuters

Mistral AI completed a new funding round led by South Korea’s Samsung 005930 -0.19 %decrease; down pointing triangle Electronics that lifted its valuation above $24 billion, giving it additional firepower to invest in model development and grow its customer base.

The French artificial-intelligence startup said it had raised 3 billion euros, equivalent to $3.48 billion, from a pool of existing investors like Samsung and Dutch chip-equipment giant ASML Holding ASML -0.31 %decrease; down pointing triangle as well as new backers such as Advent, funds and accounts managed by BlackRock and the Grand Duchy of Luxembourg.

The announcement comes about a year after Mistral AI closed an ASML-led round that gave the company an 11.7 billion-euro valuation. The latest round, co-led by Samsung, EQT-managed Scaleup Europe Fund and existing investor PSG Equity, brings its valuation above 21 billion euros, or $24.39 billion.

Johan Bergqvist, who joined Mistral AI earlier this year as chief financial officer, said in an interview that this was the largest equity fundraising round ever completed by a European tech company, and that support from ASML and Samsung showed how big tech groups continued to see promise in Mistral AI about three years after its launch.

“That’s a reflection of the growing confidence in what we can offer,” he said, noting that Mistral AI would leverage the freshly raised capital to fund a significant expansion of its models, for which the company sees growing appetite.

Mistral AI has been a strong supporter of open-weight AI models—those that are publicly available for anyone to download and modify—saying they give users the flexibility to tailor them to their needs while protecting their data and intellectual property.

Mistral AI Exceeds $24 Billion Valuation After Samsung-Led Investment Round

The Arm Mali G2-Ultra NX introduces groundbreaking capabilities for mobile graphics by incorporating dedicated neural accelerators, a state-of-the-art execution engine, and a new third-generation ray tracing unit. These advancements aim to empower developers in crafting desktop-quality gaming experiences on mobile platforms.

As mobile users seek greater visual fidelity, smoother frame rates, and immersive gameplay, developers face the challenge of meeting these escalating demands while adhering to the stringent power, thermal, and bandwidth limitations of mobile devices. Traditional rendering techniques excel at delivering high image quality; however, achieving desktop-level visual brilliance within the confines of mobile technology requires enhanced efficiency. This is where neural graphics comes into play, leveraging AI to enhance detail, create intermediate frames, and refine images—all while lessening the GPU load and overall system resource usage.

To effectively integrate neural graphics into mainstream mobile experiences, there must be a robust connection between the GPU architecture and the developer ecosystem. With over 14 billion Mali GPUs successfully deployed, Arm is well-positioned to offer these advanced features across the next generation of mobile devices.

At the core of the Arm Mali G2-Ultra NX is the first AI-native Mali GPU, featuring neural acceleration embedded directly into the graphics pipeline. This integration enables developers to produce more engaging and responsive visual experiences while adhering to mobile power constraints, expediting the adoption of neural graphics in real-world applications.

Designed specifically for neural graphics, the Mali G2-Ultra NX tightly couples neural accelerators with shader cores, allowing for simultaneous processing of both neural graphics and traditional graphics workloads. This synergy optimizes GPU memory usage and minimizes data transfer, greatly enhancing overall efficiency.

The Mali G2-Ultra NX combines its dedicated neural acceleration with an advanced execution engine and third-generation ray tracing capabilities to facilitate AI-driven graphics on mobile devices.

Three key neural technologies exemplify the GPU’s capabilities: 

1. **Neural Super Sampling (NSS)** enhances lower-resolution images into higher resolutions, dramatically decreasing the conventional rendering requirements for achieving quality visuals.
   
2. **Neural Frame Rate Upscaling (NFRU)** creates intermediate frames that help boost frame rates, resulting in more fluid gameplay.

3. **Neural Super Sampling and Denoising (NSSD)** fuses neural upscaling with denoising technologies to elevate image quality in complex, ray-traced environments where lighting and shadows are particularly intricate.

Through a collaboration between Arm and Sumo Digital, a demo called Neural Dawn illustrates how developers can embed NFRU and NSSD into their game development pipelines. This integration results in up to four times greater performance efficiency and as much as 70% reduction in memory demands compared to traditional rendering methods, facilitating the use of sophisticated graphical techniques like Unreal Engine MegaLights on mobile devices.

In addition to neural acceleration, the Mali G2-Ultra NX debuts a new execution engine—the most significant update to Mali's GPU instruction set in seven generations. This enhancement accommodates more complex graphics workloads and boosts performance, delivering up to twice the number of registers per processing cycle. This increase allows for the handling of intricate scenes more effectively, giving developers more flexibility to enrich their visuals while ensuring consistent frame rates.

Architectural improvements yield up to a 24% increase in benchmark performance and a 14% rise in non-AI gaming performance in comparison to its predecessor. These upgrades, when combined with neural acceleration via NFRU, empower longer gaming sessions at frame rates as high as 120 FPS, resulting in smoother gameplay experiences.

Moreover, Mali G2-Ultra NX features Arm's latest third-generation ray tracing unit, enabling advanced lighting, shadows, reflections, and complex geometries even on mobile devices. This new GPU architecture also reduces DRAM traffic by up to 13% during ray tracing tasks, alleviating memory pressures in demanding scenarios.

The GPU’s support for Opacity Micromaps streamlines the processing of detailed transparent geometries, enhancing the presence of elements like foliage, fabrics, and layered surfaces in mobile games. In one demonstration, this feature boosted frame rates by 30% while decreasing the workload associated with ray tracing by up to 70%, thus making intricate scene rendering more feasible on mobile.

By combining innovations in ray tracing with neural technologies, developers can achieve richer visual details without compromising performance due to mobile rendering limits. 

However, to harness the full potential of this new hardware, developers require access to familiar tools and workflows. Recognizing this, Arm invested significant time in developing its software ecosystem for neural graphics even before the launch of Mali G2-Ultra NX. This included the introduction of Arm Neural Technology and an open development kit, providing content creators and game engine developers early access to neural implementations. Notable collaborations, such as those with Tencent Games Central Tech's Magic Dawn and Unity China’s Tuanjie Engine, have facilitated the integration of these technologies into conventional development processes.

By integrating Arm Neural Technologies into their platforms, companies like Tencent Games and Unity China are paving the way for easier adoption of neural graphics, streamlining the transition from technology development to real-world gaming applications.

The Arm Neural Graphics Development Kit offers developers essential resources to transition their work into production with established graphics APIs and game engines. This comprehensive toolset includes NSS and NFRU, machine learning enhancements for Vulkan, as well as plugins compatible with prominent engines like Unreal Engine. Profiling tools, graphics optimization resources, and modeling aids allow developers to meticulously assess and enhance their neural graphics implementations.

This array of resources creates a seamless path from initial experimentation to full production, alleviating the need for studios to establish separate research frameworks or drastically alter their established workflows.

Companies like Where Winds Meet are already realizing the benefits of collaboration with Arm to incorporate Arm Neural Technology into their games, marking a significant leap from theoretical integration to actual deployment, thus setting the stage for broader adoption of neural graphics throughout the gaming landscape.

In alignment with the requirements for advanced mobile graphics, the Mali G2-Ultra NX and Arm's CSS for Mobile 2 necessitate a holistic approach to optimization, focusing on the entire system rather than solely on the GPU. This synergy between AI-driven graphics and improvements in compute efficiency, memory management, and software design equips silicon partners to handle increasingly sophisticated workloads, all while remaining within the constraints imposed by mobile technology. 

With a well-prepared ecosystem ready to implement these capabilities, the prospects for advancing mobile graphics are considerable, promising richer visual experiences and heightened performance for billions of mobile users around the world.

Exploring the Arm Mali G2-Ultra NX GPU: Providing desktop-quality mobile gaming with AI-driven graphics

A sleek, high-end handheld gaming console sits open on a wooden desk bathed in warm, natural morning light. Its internal components are exposed, with the Mali G2-Ultra NX GPU chip front and center—etched boldly with the Arm logo. From the chip springs a vibrant miniature 3D diorama of a fantasy game world (towering castle turrets, dynamic shadows, lush textures) floating just above the device, illustrating desktop-class mobile gameplay with AI-native graphics.

Inside the Arm Mali G2-Ultra NX GPU: Delivering desktop-class mobile gameplay with AI-native graphics - Arm Newsroom

- Arm Mali G2-Ultra NX integrates dedicated neural accelerators for enhanced mobile gaming graphics.  
- Features like Neural Super Sampling and Frame Rate Upscaling improve efficiency and visuals.  
- New execution engine and ray tracing hardware enhance performance and maintain smoother gameplay.  

Arm Mali G2-Ultra NX brings dedicated neural accelerators, a new execution engine and a third-generation ray tracing unit to help developers create desktop-class gaming experiences on mobile devices.

Demands from mobile users continue to grow. They want richer visuals, smoother frame rates, and increasingly immersive experiences. For developers the challenge is complex, as they aim to meet these expectations while staying within the strict power, thermal, and bandwidth limits of mobile devices.

Traditional rendering can deliver high image quality, but reaching desktop-class visual fidelity within mobile constraints demands greater efficiency. Neural graphics offers another path, using AI to reconstruct detail, generate intermediate frames, and refine images while reducing GPU demands and overall system workload.

However, bringing neural graphics into mainstream mobile experiences requires deep integration across the GPU architecture and developer ecosystem. With more than 14 billion Mali GPUs shipped to date, Arm has the scale and experience to make these capabilities accessible across the next generation of mobile devices.

That shift is at the heart of the Arm Mali G2-Ultra NX GPU. As the first AI-native Mali GPU, it integrates neural acceleration directly into the graphics pipeline, helping developers deliver richer, more responsive visual experiences within the mobile power envelope and bring neural graphics into production faster.

An AI-native GPU for neural graphics

Mali G2-Ultra NX is built for neural graphics from the ground up, with neural accelerators tightly integrated into the shader cores, allowing neural graphics workloads to run alongside graphics and compute workloads. The tight integration reuses the GPU memory system, coherent caches, and control structures, helping reduce data movement and improve efficiency across the pipeline.

Mali G2-Ultra NX combines dedicated neural acceleration, a new execution engine and third-generation ray tracing to enable AI-native graphics on mobile.

Three neural technologies that are part of Mali G2-Ultra NX demonstrate what this enables:

Neural Super Sampling (NSS) reconstructs higher-resolution images from lower-resolution renders, reducing the amount of conventional rendering work needed to produce a high-quality final image.

reconstructs higher-resolution images from lower-resolution renders, reducing the amount of conventional rendering work needed to produce a high-quality final image. Neural Frame Rate Upscaling (NFRU) generates intermediate frames to increase frame rates and deliver smoother gameplay.

generates intermediate frames to increase frame rates and deliver smoother gameplay. Neural Super Sampling and Denoising (NSSD) combines neural upscaling with denoising to improve image quality in demanding ray-traced scenes with complex lighting and shadows.

NSS reconstructs higher-resolution images from lower-resolution inputs, combining temporal data and integrated anti-aliasing to improve image quality efficiently on mobile.

NFRU uses motion, depth and rendered-frame data to generate high-quality intermediate frames for smoother mobile gameplay.

Neural Dawn, developed by Arm and Sumo Digital, showcases how developers can integrate NFRU and NSSD into a production game pipeline, delivering up to 4x higher performance efficiency and up to 70 percent lower external memory traffic compared with native rendering. These gains can make advanced techniques such as Unreal Engine MegaLights practical on mobile devices.

Explaining how NFRU and NSSD work together in the Neural Dawn game to improve mobile graphics.

By combining traditional rendering and neural graphics, developers can preserve image quality while reducing GPU and system-level workloads. For players, that means sharper images, smoother frame rates, richer scenes, and more responsive gameplay within the constraints of a mobile device.

A new execution engine increases the graphics budget

Alongside neural acceleration, Mali G2-Ultra NX introduces a new execution engine – the biggest Mali GPU instruction set architecture upgrade in seven generations – which is designed to handle increasingly complex graphics workloads while improving performance. It provides up to 2x more registers per wrap, helping the GPU handle increasingly demanding scenes more efficiently. This gives developers greater headroom for richer visuals and more complex effects, while maintaining consistent frame rates.

Mali G2-Ultra NX integrates dedicated neural acceleration directly within the shader core, sharing GPU memory and control structures to improve efficiency for neural graphics workloads.

The architectural improvements translate into up to 24 percent higher benchmark performance and 14 percent higher non-AI gaming performance compared with the previous generation. Combining the architectural improvements with neural acceleration through NFRU supports longer mobile gaming sessions at up to 120 FPS for smoother, more responsive gameplay.

A side-by-side comparison showing how the higher frame rate with NFRU improves visual smoothness on Mali G2-Ultra NX.

Third-generation ray tracing unit delivers richer lighting and shadows more efficiently

Mali G2-Ultra NX introduces Arm’s third generation of hardware ray tracing, supporting more complex lighting, shadows, reflections and geometry on mobile devices. The new GPU architecture delivers up to 13 percent lower DRAM traffic on leading ray tracing benchmarks, reducing memory pressure as scene complexity increases.

Meanwhile, hardware support for Opacity Micromaps helps handle intricate transparent geometry more efficiently, enabling richer foliage, more detailed fabrics, layered surfaces, and other complex scene elements on mobile devices. In one Arm demo, using Opacity Micromaps increased frame rates by 30 percent and reduced the ray tracing workload by up to 70 percent, demonstrating how the technology can make complex geometry more practical on mobile.

Real-time ray-traced shadows enabled by Opacity Micromaps to create a vivid environment in Moku’s Central Garden while maintaining performance.

Neural graphics complement these advances in ray tracing. Combining new ray tracing hardware with neural technologies, such as NFRU, helps deliver richer visual detail more efficiently within mobile rendering constraints.

Giving developers a practical path to bring neural graphics to production faster

However, new graphics hardware only delivers value when developers can access it through familiar tools and workflows. Arm started preparing the software ecosystem two years before Mali G2-Ultra NX with Arm Neural Technology and an open development kit for neural graphics. These gave content developers, tools developers, and game engine providers, including Tencent Games Central Tech’s Magic Dawn and Unity China’s Tuanjie Engine, early access to neural techniques, so they could begin integrating them into existing workflows before the hardware arrived.

“Neural graphics are opening up new possibilities for how we build the next generation of mobile games. Together with Arm, we are integrating Arm Neural Technologies into MagicDawn and jointly developing an NSSD technology demo with Arena Breakout Infinite. By connecting technology innovation with engines, developers and real game content, we are helping build an ecosystem that can bring more advanced graphics experiences to mobile players at scale.” Nathan Chen, Head of Technology, Tencent Games.

“Our collaboration with Arm is about more than bringing new technology into the engine — it’s about enabling the broader developer ecosystem to put that technology to work. By natively integrating Arm Neural Technology into Tuanjie Engine, we are helping developers more easily adopt advanced neural graphics and accelerate the path from technology innovation to real-world gaming across China’s gaming ecosystem.” Junbo Zhang, CEO, Unity China.

Tencent Games and Unity China explain how they are bringing neural graphics into familiar development workflows and helping move AI-native mobile graphics from technology to production.

Today, the Arm Neural Graphics Development Kit provides the software and tools needed to move that work toward production across established graphics APIs, game engines, and custom workflows. The developer stack includes NSS and NFRU, along with machine learning extensions for Vulkan and plug-ins for leading game engines, including Unreal Engine. An SDK supports custom-engine integration, while profiling, graphics optimization, training, and model-optimization tools help developers evaluate and fine-tune neural graphics workloads.

All these resources provide a path from experimentation to production without requiring studios to build bespoke research infrastructure or change the development workflows they already use.

As Keli Zhou, Engine Lead for Where Winds Meet, says, “Through our collaboration with Arm, we are integrating Arm Neural Technology into the Messiah Engine and bringing it into real-world game development. Where Winds Meet will be among the first games to bring Arm Neural Technology to players, marking an important step from engine-level integration to real-world deployment. Together, we are paving the way for broader developer adoption of neural graphics across the gaming ecosystem.”

The AI-native graphics foundation for Arm CSS for Mobile 2

Advanced mobile graphics require optimization across the entire system, not just the GPU. Within Arm CSS for Mobile 2, Mali G2-Ultra NX brings AI-native graphics together with compute, memory, software and implementation optimizations to help silicon partners manage increasingly demanding workloads within mobile power and thermal constraints.

Combined with an ecosystem ready to put these capabilities into production, this raises the ceiling for mobile graphics, enabling richer visuals, and smoother experiences delivered to billions of mobile devices.

Inside the Arm Mali G2-Ultra NX GPU: Delivering desktop-class mobile gameplay with AI-native graphics

Thanks to advancements in artificial intelligence, robots are capable of visual perception, learning, and being reconfigured for various tasks. In the realm of service industries, for instance, a robot initially designed for making beds can adapt its skills to perform additional duties, such as organizing towels, emptying waste bins, or undertaking any other assigned task. This versatility enhances their functionality and efficiency, making them invaluable assets in numerous applications.

UK's top tech leader claims chip shortage is delaying AI cancer treatments.

A dim, moody biotech lab bathed in warm tungsten light—empty vials stamped “AI Cancer Cure” lie abandoned on a metal table as tiny microchips spill across its surface. In the shadows, a lone corporate executive stands with their back to us, facing idle servers and a Union Jack flag hanging on the wall. The scene conveys stalled progress and mounting frustration.

AI cancer cures slowed by chip shortage, says UK's biggest tech boss - BBC

- AI-equipped robots can see, learn, and be reprogrammed for various tasks.  
- A robot can switch from making beds to arranging towels or cleaning dustbins.  
- These advancements significantly enhance flexibility and efficiency in the service industry.  

"With artificial intelligence, these robots can see, learn, and essentially be reprogrammed for new tasks. So, in the service industry, the robot that was programmed to make a bed can also learn how to arrange the towels in a room, or clean the dustbins, or whatever you want to go off and do."

Nebius rises as stocks in AI infrastructure gain momentum

More from AI

David Bahnsen cautions that high capital expenses pose the biggest threat to AI investors.

Google Assists Accenture in Integrating AI Engineers with Clients

According to Google, Chinese hackers are leveraging AI on compromised networks to evade detection.

OECD Reports AI Implementation Linked to Worldwide Decrease in Student Reading Abilities