How Google’s Custom AI Chip is Disrupting the Tech Industry

SEI 275896064

Ironwood is Google’s latest tensor processing unit

Nvidia’s dominance in the AI chip market is facing challenges due to a new specialized chip from Google, with several companies, such as Meta and Anthropic, planning to invest billions in Google’s tensor processing units.

What is TPU?

The growth of the AI industry heavily relies on graphics processing units (GPUs), which are designed to execute numerous parallel calculations at once, unlike the sequential processing of central processing units (CPUs) found in most computers.

Originally engineered for graphics and gaming, GPUs can handle operations involving multiple pixels simultaneously, as stated by Francesco Conti from the University of Bologna, Italy. This parallel processing is advantageous for training and executing AI models, particularly with tasks relying on matrix multiplication across extensive grids. “GPUs have proven effective due to their architecture fitting well with tasks needing high parallelism,” Conti explains.

However, their initial design for non-AI applications introduces some inefficiencies in how GPUs handle computations. Google launched Tensor Processing Units (TPUs) in 2016, which are optimized specifically for matrix multiplication, the primary operation for training and executing large-scale AI models, according to Conti.

This year, Google introduced the 7th generation TPU called Ironwood, which powers many of the company’s AI models, including Gemini and AlphaFold for protein modeling.

Are TPUs Superior to GPUs for AI?

In some ways, TPUs can be considered a specialized segment of GPUs rather than an entirely separate chip, as noted by Simon McIntosh-Smith from the University of Bristol, UK. “TPUs concentrate on GPU capabilities tailored for AI training and inference, but they still share similarities.” However, tailored design means that TPUs can enhance the efficiency of AI tasks significantly, potentially leading to savings of millions of dollars, he highlights.

Nonetheless, this focus on specialization can lead to challenges, Conti adds, as TPUs may lack flexibility for significant shifts in AI model requirements over generations. “A lack of adaptability can slow down operations, especially when data center CPUs are under heavy load,” asserts Conti.

Historically, Nvidia GPUs have enjoyed an advantage due to accessible software that assists AI developers in managing code on their chips. When TPUs were first introduced, similar support was absent. However, Conti believes that they have now reached a maturity level that allows more seamless usage. “With TPUs, we can now achieve similar functionality as with GPUs,” he states. “The ease of access is becoming increasingly crucial.”

Who Is Behind the Development of TPUs?

While Google was the first to launch TPUs, many prominent AI firms (referred to as hyperscalers) and smaller enterprises are now venturing into the development of their proprietary TPUs, including Amazon, which has created its own Trainium chips for AI training.

“Many hyperscalers are establishing their internal chip programs due to the soaring prices of GPUs, driven by demand exceeding supply, making self-designed solutions more cost-effective,” McIntosh-Smith explains.

What Will Be the TPU’s Influence on the AI Industry?

For over a decade, Google has been refining its TPUs, primarily leveraging them for its AI models. Recently, changes are noticeable as other large corporations like Meta and Anthropic are investing in considerable amounts of computing power from Google’s TPUs. “While I haven’t seen a major shift of big clients yet, it may begin to transpire as the technology matures and the supply increases,” McIntosh-Smith indicated. “The chips are now sufficiently advanced and prevalent.”

Besides providing more options for large enterprises, diversifying their options could also make economic sense, he notes. “This could lead to more favorable negotiations with Nvidia in the future,” he adds.

topic:

Source: www.newscientist.com

British Firms Poised to Seize a Major Share of the AI Chip Market

TThe UK holds a unique and advantageous position to contribute significantly in the new era of artificial intelligence, provided it seizes the chance to establish the production of millions of computer chips, an area that is often misunderstood.

AI technology demands a vast quantity of chips, and a collaborative national initiative could fulfill up to 5% of the global requirement.

Our legacy in chip design is unparalleled, beginning with the first general-purpose electronic computer, the initial electronic memory, and the first parallel computing system. Presently, Arm, based in Cambridge, is a prominent player that designs over 90% of the chips found in smartphones and tablets worldwide.

Given this background, it is certainly plausible that British companies can capture a notable share of the AI chip market. A target of 5% is both conservative and achievable. Our distinguished universities, a flourishing foundational AI company like DeepMind, and a strong innovation ecosystem equip the UK with the tangible resources necessary to compete.

The potential gains are tremendous. The global market for AI chips is expected to soar to $700 billion (£620 billion) annually by 2033, surpassing the entire current semiconductor market. Achieving that 5% share would translate to an influx of $35 billion in new revenue and the creation of thousands of high-paying jobs.

AI is set to transform not only the economy but also societal structures and security. Unfortunately, many do not grasp where its true value and strategic influence lie.

In this contemporary gold rush, real wealth is accessible not only to those mining digital gold but also to those who provide the tools for the task. I witnessed this firsthand from 1997 to 2006 when Gordon Moore and Andy Grove helped establish Intel’s board and founded the company in California. They set the groundwork for the first technology revolution, much like Nvidia is doing today on an even larger scale.

UK engineers, intellects, businesses, and investors excel in this domain. However, government collaboration is crucial.

While consumers are captivated by the generative marvels of OpenAI, the true market winner is Nvidia, the entity that provides the advanced chips facilitating such achievements. OpenAI’s estimated value stands at merely 1/10th that of Nvidia. AMD, a semiconductor design company, holds a distant second place, while emerging firms like Cerebras and TenTrent strive for a share of the market.

All AI models and applications, ranging from autonomous robots to real-time translation services, depend heavily on advancements in chip technology. Chips are the new oil of the digital economy, dictating the speed and efficiency with which future applications can be developed. Currently, the only major players in the AI field seeing true profitability are chip manufacturers.

Concerns have arisen that China may commoditize AI chips similar to its approach with solar technology, leading to dramatic price fluctuations and undercutting existing companies. The situation is more complex. U.S. export controls will restrict China’s access to advanced chip manufacturing technology for the next decade, significantly curtailing its capacity to dominate the high-end AI chip arena. This reality positions the U.S. as a key player and creates a substantial opportunity for its closest ally, the UK, which excels in chip design.

The UK has already birthed several companies in this sector, such as Fractile, Flux, and Oriole. However, we lack the necessary scale to capitalize on the opportunity. Instead of competing with Nvidia in data center computing, we should focus on specialized applications that usher in innovation, like robotics, factory automation, medical devices, and autonomous vehicles.

These domains offer ample opportunities for inventive architectures and new competition.

Skip past newsletter promotions

Too frequently, Britain’s industrial strategy is impeded by national insecurity and a lack of confidence. This must change. Primarily, governments must advocate decisively for our intent to excel in AI chips.

Secondly, we should aim to double our chip design workforce from the current 12,000 within a decade and encourage more talented individuals to pursue electrical engineering and computer science through generous scholarships. A target of 1,500 new students each year is achievable. Universities must offer relevant courses, and governments need to enhance financial support.

Thirdly, the UK should fully utilize its investment instruments: the Sovereign AI Fund, the British Business Bank, the National Wealth Fund, and the Ministry of Defence’s initiatives to ‘buy British’.

Fourthly, the UK-US strategic partnership must serve as a foundation for greater collaboration with leading US chip manufacturers and facilitate access to their state-of-the-art sub-3 nanometer manufacturing technologies. Collaborating with our U.S. partners to develop a robust supply chain and innovation pipeline is essential.

If the UK commits fully, the emerging age of AI could be characterized not only by code but also by silicon, leaving a distinctly British legacy.

Source: www.theguardian.com

Nexperia Halts Chip Supply to China Amid Global Automotive Production Concerns

Nexperia, the automotive semiconductor manufacturer based in the EU and at the heart of the geopolitical tensions, has stopped all supplies to its factories in China, intensifying a trade war that risks shuttering production for automakers globally.

This week, the company communicated with its clients about the suspension of all supplies to its Chinese facility.

In September, the Netherlands utilized national security legislation to take control of the semiconductor maker due to fears that its Chinese owner, Wingtech Technologies, intended to transfer intellectual property to another affiliated company. The Dutch authorities stated: Threatened the future of Europe’s chip production capacity and subsequently dismissed Wingtech Chairman Zhang Xuezheng as CEO.

In retaliation, China ceased exports from all Nexperia factories and warned that this embargo could lead to the closure of production lines at EU car manufacturers within days.

The continuing lockdown jeopardizes the supply chain, as numerous Nexperia products produced in Europe—including wafers used to manufacture chips—were typically sent to factories in China for packaging and distribution.

Nexperia’s interim CEO, Stephen Tilger, stated on Sunday that shipments to its Dongguan factory in Guangdong province have been halted due to a “direct result of local management’s recent failure to comply with agreed contractual payment terms,” according to excerpts first released by Reuters.

Nexperia remains optimistic about resuming shipments and is hoping to de-escalate the situation. A source familiar with the developments indicated that shipments might recommence once contractual payments are made. Additionally, the company will continue sending products to its Malaysian facility, which is smaller than the Chinese one.

Automakers are expressing concerns over potential disruptions caused by shortages of crucial components essential for modern vehicles.

The automotive sector faced severe semiconductor shortages following the coronavirus pandemic, but it was not Nexperia’s lower-cost power control chips that were impacted—it was more advanced chips. The company usually ships over 100 billion items annually, utilized in various applications from airbags and adjustable seating to side mirrors and central locking.

Nissan Motor Co. announced this week that it has sufficient chips to last until early November, while competitor Honda reported halting production at its Mexican facility. Mercedes-Benz described its situation as “manageable” in the short term, yet is exploring alternatives. Volkswagen suggested that its annual profit goals could be compromised without adequate chip supply.

Conversely, Toyota, the world’s largest automaker, informed reporters at an auto show in Tokyo on Friday that it is not experiencing significant supply challenges, even though production might ultimately be affected.

EU trade commissioner Maroš Šefčović aims to initiate further discussions with Chinese officials following meetings in Brussels with both Chinese and EU representatives to address the export ban on Nexperia and restrictions on rare earth minerals supply.

Additionally, on the same day, the bloc’s technical commissioner Hena Virkunen met with Nexperia’s interim leader after discussions with European chip manufacturers Infineon, ST, and NXP the previous day.

Post-meeting, he noted that discussions with Nexperia underscored the EU’s necessity for new tipping laws as three lessons identified from the ongoing crisis: enhanced visibility of chip inventory in the pipeline, the importance of investing in chip supply despite costs, and the need for reserve inventories.

Skip past newsletter promotions

“Diversifying stockpiles and supplies is crucial to our collective resilience,” she stated.

The German Automotive Industry Association (VDA) expressed concern on Thursday that without a swift resolution to the situation at Nexperia, it could lead to “significant production restrictions and even suspensions in the near future.”


Businesses in the UK are likely to be impacted as well. Nexperia manufactures some of its chip wafers at a plant established by Dutch company Philips in Manchester.

Previously, Nexperia owned another factory in south Wales but was blocked by the UK government from completing its acquisition of the Newport wafer factory due to national security concerns, given its ultimate Chinese ownership. Subsequently, US semiconductor firm Vishay Intertechnology acquired the factory in November 2023.

Wingtech has yet to respond to requests for comments.

Source: www.theguardian.com

Record-Breaking Chip Defies Moore’s Law by Expanding Vertically

Stacking semiconductor transistors could aid in overcoming Moore’s law

Kaust

As semiconductor manufacturers make their products smaller, they encounter limitations on the computing power that can be integrated into a single chip. A groundbreaking chip may offer a solution to this dilemma and advance the creation of sustainable electronics.

Since the 1960s, enhancing electronic capabilities has revolved around miniaturizing their fundamental components, transistors, and packing them more densely onto chips. This trend was encapsulated by Moore’s Law, which posited that the number of components on a microchip doubles every year. However, this phenomenon began to falter around 2010. Li Xiaohan and colleagues at Saudi Arabia’s King Abdullah University of Science and Technology have suggested that the answer to this challenge might be to build upwards instead of inwards.

They engineered a chip featuring 41 vertical layers of two distinct semiconductor types, separated by insulating material. This stack of transistors is approximately ten times taller than any previously created. To evaluate its efficiency, the team produced 600 duplicates, all demonstrating consistent performance. Some of these stacked chips were utilized to execute various fundamental operations required by computers or sensing devices, showing performance levels comparable to traditional non-stacked counterparts.

Li mentions that producing these stacks necessitates a manufacturing method that requires less energy compared to standard chip production. Team members, including Thomas Anthopoulos from the University of Manchester in the UK indicates that while the new chip may not lead to advanced supercomputers, its application in everyday devices like smart home gadgets and wearable health monitors could significantly lower the carbon footprint of the electronics industry while enhancing functionality with each additional layer.

How high will the stack rise? “The possibilities are endless; we can keep pushing the limits. It’s just a journey of determination,” Anthopoulos states.

However, he notes that engineering hurdles persist regarding the temperature tolerance of the chip before it fails. Muhammad Alam from Purdue University in Indiana comments that it’s analogous to trying to keep cool by layering on multiple hoodies; each additional layer raises the heat. Alam asserts that the chip’s current thermal threshold of 50 degrees Celsius would need to rise by over 30 degrees Celsius to become practical for real-world application. Nonetheless, he believes that for electronics to progress in the near future, pursuing vertical growth is the only viable strategy.

topic:

Source: www.newscientist.com

OpenAI Secures Billion-Dollar Chip Partnership with AMD Technology

On Monday, OpenAI and semiconductor manufacturer AMD revealed that they have entered into a multi-billion dollar agreement concerning chips, which will allow the creators of ChatGPT to purchase significant equity stakes in the chipmaker.

This arrangement provides OpenAI the chance to acquire 10% of AMD, reflecting substantial confidence in the company’s AI chips and software. Following the announcement, AMD’s stock soared by over 30%, contributing approximately $800 billion to its market capitalization.

“We are excited to announce our dedication to delivering a variety of services to our clientele,” stated Forest Norod, AMD’s Executive Vice President.

These recent investment commitments underscore OpenAI’s significance, as the increasing demands of the AI sector drive companies to advance AI technologies that rival or surpass human intelligence. OpenAI’s CEO, Sam Altman, pointed out that the primary limitation on the company’s expansion is access to computing resources, particularly extensive data centers equipped with advanced semiconductor chips. Last week, Nvidia declared a $100 billion investment in OpenAI, further solidifying the collaboration between these leading AI firms.


The agreement announced on Monday encompasses the deployment of hundreds of thousands of AMD AI chips or graphics processing units (GPUs) totaling 6 gigawatts over several years, starting in the latter half of 2026. AMD confirmed that OpenAI will establish a 1 Gigawatt facility utilizing the MI450 series chips beginning next year.

Additionally, AMD issued a warrant that enables OpenAI to purchase up to 160 million shares of AMD at just one cent each during the course of the chip trading.

AMD’s executives anticipate that this transaction will result in tens of millions of dollars in annual revenue. Due to the expected ripple effects of this contract, AMD has projected more than $100 million in new revenue over four years from OpenAI and other clientele.

“This marks a trailblazing initiative in an industry poised to significantly influence broader ecosystems, attracting others to join,” remarked Matt Hein, AMD’s Head of Strategy.

Skip past newsletter promotions

This agreement with AMD is expected to significantly bolster OpenAI’s infrastructure to fulfill its operational requirements, Altman confirmed in a statement.

However, it remains unclear how OpenAI plans to finance this substantial deal with AMD. According to media reports, the deal is estimated to be worth around $500 million, yielding approximately $4.3 billion in revenue in the first half of 2025 while burning through $2.5 billion in cash.

Source: www.theguardian.com

Nvidia and AMD Allegedly Set to Contribute 15% of China’s Chip Sales Revenue to the US

Nvidia and AMD have made a groundbreaking agreement to allocate 15% of their revenue from chip sales in China to the US government, a deal aimed at securing a semiconductor export license. The Financial Times reported on Sunday.

This revenue-sharing initiative includes Nvidia’s H20 chips and AMD’s Mi308 chips, with details emerging from US officials indicating that the Trump administration is yet to determine the allocation of these funds.

An anonymous official stated that the chipmakers consented to this Quid Pro Quo arrangement as a prerequisite for obtaining a Chinese export license last week.


According to export management specialists, this marks the first time US companies have agreed to a revenue-sharing model in exchange for export licenses, as reported by the newspaper. Donald Trump has reportedly encouraged these firms to invest in the US to “offset” the tariffs imposed.

In a statement to Reuters, an Nvidia spokesperson mentioned, “We haven’t shipped H20 to China for months, but we are optimistic that export control regulations will enable us to compete globally.”

AMD did not provide an immediate response to inquiries for comment.

Last week, the US Department of Commerce commenced the issuance of licenses to NVIDIA for the export of H20 chips to China, removing a significant barrier to entering key markets.

In July, the US overturned an earlier ban on the sale of H20 chips to China. Nvidia had specifically modified its microprocessors for the Chinese market to align with the Biden administration’s AI chip export regulations.

Nvidia’s chips are pivotal in driving the current AI surge, and the company became the first to surpass a market valuation of $4 trillion in July.

However, Nvidia faces growing scrutiny from Chinese regulatory bodies, with challenges likely to persist. Recently, China’s Cyberspace Watchdog summoned Nvidia to clarify concerns regarding a potential “backdoor” security risk that might grant remote access or control over the chip. Nvidia refuted these claims.

Nonetheless, concerns have been echoed in Chinese state media. Earlier this month, it was reported that officials stated Nvidia needs to furnish “persuasive security proofs” to assuage worries over security risks for Chinese users and regain trust in the market. Additionally, the WeChat national media account highlighted potential security risks posed by the H20 chip, suggesting the possibility of “remote shutdown” features via a hardware “backdoor.” Nvidia has yet to respond to these allegations.

Reuters and

Source: www.theguardian.com

Nvidia CEO: US Chip Export Controls Misfire by Boosting China’s Progress

Jensen Huang, head of Nvidia, stated that US chip export controls are a “fail” during his remarks at the High-Tech Forum on Wednesday.

In an effort to limit China’s military advancements and maintain US dominance in the AI sector, successive US administrations have placed restrictions on the sale of advanced AI chips to China. However, Huang indicated at the Computex Tech forum in Taipei that these controls have inadvertently motivated Chinese developers.

“The local companies are exceptionally skilled and highly motivated, and export control has provided them with the momentum, energy, and governmental backing to speed up their progress,” Huang shared at the Computex Tech Show in Taipei.

“On the whole, I believe export control has been a failure.”



“It’s crucial to acknowledge that China boasts a dynamic technological ecosystem, with 50% of the world’s AI researchers, and excels in software development,” Huang emphasized.

Nvidia, known for its high-performance GPUs, faces challenges due to US chip export regulations. Huang mentioned on Wednesday that the company has incurred “billions of dollars” in losses, with its share of the AI chip market in China plummeting from nearly 95% to 50% since the Biden administration took office.

According to the Financial Times, Huang made an unannounced trip to Beijing last month.

This visit took place shortly after new US restrictions prohibited the shipment of Nvidia’s H20 Datacentre GPU to China.

The US government informed NVIDIA that the new regulations aim to mitigate the risk of the product being “used in Chinese supercomputers.”

Huang’s Beijing conference reportedly focused on the AI company’s latest chip design, as per the FT report.

Last week, the Trump administration rolled back certain existing controls on chip sales to China after several countries expressed that they were being excluded from the essential technologies required for AI development.

Nonetheless, they issued new guidelines for other nations, warning that utilizing high-tech AI semiconductors produced in China, especially chips from Huawei, could breach existing US export regulations.

In retaliation, China accused the United States of “misusing export controls to suppress and restrict China.” The Commerce Department stated on Wednesday that the warning exemplifies “unilateral bullying and protectionism, significantly jeopardizing the stability of the global semiconductor industry and supply chains.”

Moreover, it cautioned that organizations or individuals enforcing or supporting such actions might be violating Chinese law.

Source: www.theguardian.com

Nvidia CEO Makes Surprise Visit to Beijing Following Chip Sales Restrictions in China

The CEO of American chip maker Nvidia recently visited Beijing shortly after the US imposed new restrictions on the sale of AI chips to China.

According to state media-affiliated social media accounts, Jensen Fan’s unexpected visit was in response to an invitation from a trade agency.

China Central Television reported that Huang met with Ren Hongbin, the head of China’s Council to promote international trade.

The official English outlet of the Communist Party released a photo of Huang in Beijing, stating, “It’s three months since I promised to continue working with #China during my last visit.” The hashtag #OpportunityChina was included, previously used in a post promoting US-China exports.

This visit comes amidst a turbulent week for Nvidia. The recently announced US restrictions affect the shipment of the H20 DataCentre GPU, a specialized low-power version of Nvidia chips designed to comply with restrictions on sales to China under the Biden administration.

Amidst the ongoing race for AI dominance between the US and China, the US government informed Nvidia that the new rules aim to mitigate the risk of its products being utilized in Chinese supercomputers.

The company estimates that these new restrictions will cost around $5.5 billion (£4.2 billion) and experienced a 7% drop in its shares on Wednesday.

The tech industry has been under pressure due to US restrictions on high-tech supply to China and widespread tariffs. Nvidia’s shares decline is part of a broader trend in the sector which has seen many companies experiencing significant drops in recent weeks. Trump’s threats of separate tariffs on the global semiconductor industry further add to the uncertainty.

Following the announcement of the new Nvidia chip restrictions, semiconductor companies have pledged to invest up to $500 million in AI infrastructure in the US over the next four years.

Nvidia designs chips but outsources production to contractors like Taiwanese semiconductor manufacturers. TSMC, for instance, has committed to large-scale investment projects in the US, exempting them from tariffs. In response, the White House attributed Nvidia’s decision to “the Trump effect.”

Reportedly, Huang also met with Liang Wenfeng, the founder of Deepseek in Beijing, to discuss new chip designs for AI companies that would not trigger another US ban. Deepseek gained attention in January for its advanced AI chatbot developed with minimal investment, shaking up the tech industry and impacting global stock markets.

The US House of Representatives China Committee has raised concerns about Deepseek potentially using an export-controlled chip to power its AI app, posing a national security threat.

Huang has publicly stated that Nvidia is committed to advancing AI globally while complying with legal requirements and technological advancements under the Trump administration. He reassured reporters that the company will continue its progress in the field.

Skip past newsletter promotions

Huang’s visit to Beijing created a buzz on social media in China and Taiwan. As a Taiwanese celebrity, he was welcomed by a large number of fans on his recent visit, generating excitement and reports about his activities.

The chaos caused by Trump’s tariffs has raised concerns among global markets and governments, including US allies. Amidst changing tariff rates and negotiations, the focus remains on reshaping trade agreements to address trade imbalances and economic concerns.

Trump’s recent talks with Japan indicate a strategic approach to trade negotiations with various countries, signaling a priority for the US administration in reshaping global trade relations.

Additional report by Jason Tzu Kuan Lu

Source: www.theguardian.com

Nvidia’s finances to take a $5.5 billion hit amid US restrictions on AI chip exports to China.

Nvidia has announced that it is expecting a $5.5 billion (£4.1 billion) impact following the ban imposed by Donald Trump’s administration on chip designers selling crucial artificial intelligence chips in China.

In an official statement released late Tuesday, the company disclosed that the H20 AI chip, specifically tailored for the Chinese market to comply with export regulations, will now require a special license for sale in China indefinitely.

The US government, engaged in a competition with China for AI supremacy, informed Nvidia that new regulations have been enacted to mitigate the risk of their products being utilized in Chinese supercomputers.

As a result, the chip manufacturer is set to incur $5.5 billion in losses for the financial quarter ending on April 27th due to its investment in H20 chips.

Nvidia, known for driving significant advancements in AI technology, has delivered substantial returns for investors, with its stock surging over 1,400% since 2020, making it one of the few trillion-dollar companies in the US.

However, the news on Tuesday caused Nvidia’s stock to fall by approximately 6% in after-hours trading, potentially wiping out billions of dollars in market value by Wednesday’s opening bell.

In Asia, chipmakers like Samsung Electronics and SK Hynix from South Korea saw a 3% decline in their stocks overnight, while US competitors like senior equity microdevices dropped by 7% in after-hours trading.

Although the chip industry has been exempt from the 10% tariff that began on April 5th, Trump indicated this week that he plans to impose tariffs on imported semiconductors and mentioned that some companies in this sector may have flexibility.

The US Department of Commerce has recently launched an investigation into the impact of chip and drug imports on national security.

The US heavily relies on chip imports from Taiwan, with Trump previously imposing a 32% tariff on products from the country before suspending most tariffs last week.

Skip past newsletter promotions

Nvidia also revealed plans to invest up to $500 million in AI infrastructure in the US over the next four years to bolster its American manufacturing presence. While Nvidia designs chips, it outsources production to contractors, including Taiwanese semiconductor manufacturers.

Under the Biden administration, US officials had initially barred Nvidia and other AI chip manufacturers from selling advanced chips to China in October 2022. In response, Chinese authorities tightened controls over the tools and processors necessary for semiconductor production.

Source: www.theguardian.com

Chip tariffs bolster US AI infrastructure as Nvidia looms larger.

Nvidia, a chip designer, has announced plans to invest up to $500 million (£37.8 billion) in artificial intelligence infrastructure in the United States over the next four years. This move comes in response to President Trump’s tariffs, with signs of manufacturers shifting their investments to American businesses.

The decision follows Trump’s recent tariffs on semiconductors, which are primarily produced by Nvidia in Taiwan. The company’s CEO, Jensen Huang, visited the president at Mar-a-Lago earlier this month, prompting repeated threats from Trump. Nvidia aims to establish fully operational “supercomputer” facilities in the US and collaborate with manufacturing partners to construct factories.

Production of the Blackwell graphics processing unit has already commenced at the TSCM’C factory in Phoenix, Arizona. Additionally, new plants are being developed at Foxconn in Houston and Wistron in Dallas, with expectations of increased mass production within the next 12-15 months.

Huang emphasized that enhancing American manufacturing capabilities is crucial for meeting the growing demand for AI chips and supercomputers, thereby strengthening the company’s supply chain and resilience.

The White House hailed Nvidia’s commitment as a result of the “Trump effect.” Nvidia’s stock market value has surged over the years, driven by the demand for AI chips, but faced challenges due to tariff uncertainties resulting in a drop in stock prices.

Global markets reacted cautiously, hoping for some relief from Trump’s new taxes. While markets in Japan and South Korea rose, Hong Kong and China experienced declines. In Europe, markets showed gradual recovery, with the UK’s FTSE 100 index rising, along with Germany’s DAX and France’s CAC.

Trump’s plans to impose tariffs on semiconductor and drug imports have raised concerns. The US Department of Commerce has launched an investigation into the impact on national security. Amidst these decisions, companies like Novartis are investing in the US drug sector to mitigate tariff threats.

Source: www.theguardian.com

After securing the meta contract, Arm unveils its own chip launch

The UK semiconductor designer ARM reportedly plans to launch its own chip this year after landing Meta as one of its first customers.

The move represents a massive overhaul of the SoftBank-owned group’s business model, licensing chip blueprints to Apple and Nvidia.

ARM CEO Rene Haas is set to announce its first in-house chip as early as this summer, according to a Financial Times report citing people familiar with the plan.

Skip past newsletter promotions

Since the company was founded in 1990, more than 300 million chips have been shipped based on ARM design, and almost all world smartphones are based on ARM technology. Moving from chip design to manufacturing a full, proprietary processor could also compete with the largest customers in the £500 million semiconductor industry.

ARM declined to comment. The company’s shares rose more than 6% on Thursday after finance reported its plans.

Financial Times also has its own intellectual property by creating a vast infrastructure network for AI, which has shifted to the production of AI chips by the son of SoftBank founder Masayoshi, and building a vast infrastructure network for artificial intelligence. It reported that it is one step in a big plan to make more money from.


Masayoshi Son, CEO of Softbank Group. Photo: Mitsui/Aflo/Rex/Shutterstock

Last month, Son announced its Stargate initiative at Openai. It spends an estimated £400 million building AI infrastructure, funded by Abu Dhabi State Fund MGX and Oracle, and is armed as a major technology partner alongside Microsoft and Nvidia.

According to those familiar with the plan, ARM’s chips are the central processing units (CPUs) of servers in large data centers and are expected to be customizable for clients, including Meta. These people said production will be outsourced to manufacturers such as Taiwan Semiconductor Manufacturing Co.

Another transaction essential to ARM’s chipmaking project is SoftBank’s anticipated acquisition of Ampere. It could be valued at nearly $6.5 billion (£5.155 billion).

Cambridge Headquarted Arm has more than doubled to $173 million since it was listed on the Nasdaq in 2023. Before SoftBank took over it in 2016, ARM was previously listed in London.

Meta is the latest big tech company that looks to ARM for power-efficient server chips instead of Intel and AMD. Meanwhile, ARM’s Nvidia partnership with Amazon has driven the rapid growth of data centers that power Openai, Meta, and human AI assistants.

Source: www.theguardian.com

Chinese AI Chip Company Recently Blacklisted Over Arms Concerns Now Granted Access to UK Technology

Chinese engineers are developing artificial intelligence chips for use in “advanced weapons systems” and have been granted access to cutting-edge British technology, as reported by the Guardian.

Moore Thread and Viren Technology, described as “China’s leading AI chip designers,” have been subject to U.S. export controls for their chip development. It is noted that the technology can provide artificial intelligence capabilities for the advancement of weapons of mass destruction, advanced weapons systems, and high-tech surveillance applications that raise national security concerns.

Before being blacklisted in the US in 2023, the companies had a broad license with UK-based Imagination Technologies, known for its expertise in designing advanced microchips essential for AI systems.

Imagination Technologies, a representative of the UK technology industry, denied intentionally trying to relocate its cutting-edge secrets to China. Representatives from Imagination confirmed the existence of licenses to Moore Thread and Viren Technology.

Allegations have arisen regarding Imagination’s partnerships with Chinese companies and the potential risks of knowledge transfer. Tensions between business with China and national security concerns have been highlighted by these developments.

Since 2020, at least three Chinese companies have obtained licenses to use Imagination’s chip designs, raising concerns about the potential misuse of intellectual property.

Imagination has worked closely with Apple in the past, contributing to the development of iPhone chips. However, concerns have been raised about the risks of sharing too much of its intellectual property with Chinese companies.

The acquisition of Imagination by a Chinese-backed buyer in 2017 raised further concerns about technology transfer and national security implications.

Imagination’s arrangements with Chinese customers are considered “totally normal” and have been described as limited in scope, duration, and usage rights.

Imagination’s policy of not doing business with companies on the US government’s Entity List raises questions about the termination of licenses granted to Chinese companies in October 2023.

A new report from the UK and China transparency research institute sheds light on further questions surrounding Chinese companies’ relationships with Imagination.

Moore Thread and Biren Technology, two Chinese chipmakers, have faced scrutiny for their development of GPUs for AI systems with potential ties to Imagination’s technology.

Funding for Biren Technology comes from the Russia-China Investment Fund, sparking concerns about deepening alliances between China and Moscow in the tech industry.

Source: www.theguardian.com

Google reveals revolutionary quantum computing chip

Measuring just 4cm square, Google has developed a computing chip with unprecedented speed. In just five minutes, this chip can complete tasks that would take conventional computers 10 billion years to finish – a mind-boggling number surpassing the age of our universe.

The chip, named Willow, is the size of an After Eight Mint and could revolutionize drug development by accelerating the experimental phase. Recent advancements suggest that within five years, quantum computing will transform research and development across various industries.

Willow boasts fewer errors, enhancing the potential of artificial intelligence. Quantum computing leverages matter existing in multiple states simultaneously to make vast calculations beyond previous capabilities, expediting advancements in medicine and technology.

However, concerns remain about security vulnerabilities posed by quantum computing – the ability to breach even the most robust encryption systems.

Google Quantum AI, alongside other entities like Microsoft, Harvard University, and Quantinum, is working on harnessing quantum mechanics for computing. Overcoming challenges in error correction has paved the way for significant speed enhancements and groundbreaking developments.

Quantum processors are evolving rapidly, surpassing traditional computers and unlocking new possibilities for quantum computations. The potential for quantum computers to exist in multiple states simultaneously promises remarkable capabilities across various fields.

Skip past newsletter promotions

Dr Peter Leake, Research Fellow at the University of Oxford’s Quantum Institute and founder of Oxford Quantum Circuits, acknowledges the rapid advancements in quantum computing technology. While applauding Google’s progress in error correction, he highlights the need for practical applicability in real-world scenarios.

As quantum computing approaches practical implementation, collaboration across various fields becomes crucial to navigate challenges and harness the full potential of this groundbreaking technology.

Source: www.theguardian.com

Intensifying Chip War: New U.S. Regulations Targeting China’s Semiconductor Industry

The United States announced new export restrictions targeting China’s advanced semiconductor manufacturing capabilities, drawing immediate criticism from the Chinese government.

The U.S. government is expanding efforts to curb exports to China of cutting-edge chips that can be used in advanced weapons systems and artificial intelligence.

Monday’s announcement comes weeks before Donald Trump returns as president, where he is expected to strengthen Washington’s hawkish stance on China. Commerce Secretary Gina Raimondo said Monday that President Joe Biden’s term has been particularly challenging in “strategically addressing China’s military modernization through export controls.”

Biden’s national security adviser, Jake Sullivan, said: “The United States has taken significant steps to ensure that our technology is not used by adversaries in ways that threaten our national security.” . The U.S. government continues to work with allies and partners to “actively and aggressively protect our world-leading technology and know-how from being used to undermine our national security.”

The Chinese government pledged on Monday to protect its interests, with a spokesperson for the Chinese Ministry of Commerce saying the United States was “abusing export control measures” and “impeding normal economic and trade exchanges.”

The latest U.S. rules include restrictions on sales to 140 companies, including Chinese semiconductor companies Pyotek and SiCarrier, without additional permits. The Commerce Department said they also affect Nowra Technology Group, which makes chip manufacturing equipment. Others include entities in Japan, South Korea and Singapore.

The new U.S. rules also include regulations for 20 types of chip manufacturing equipment and three types of software tools for semiconductor development or production. “We are in constant dialogue with our allies and partners to reevaluate and update our controls,” said Alan Estevez, Undersecretary of Commerce for Industry and Security.

Netherlands-based computer chip equipment maker ASML, the only manufacturer of cutting-edge chip-making machinery, said it does not expect new U.S. regulations to impact its latest financial metrics. Ta.

ASML said the latest U.S. regulations, if implemented by the Dutch government, will impact exports of deep ultraviolet lithography (DUV) systems to some chip manufacturing plants in China. ASML is the only manufacturer of extreme ultraviolet lithography equipment (EUV) that produces cutting-edge chips. The company already cannot sell EUV equipment to China because of existing government restrictions on the use of US technology.

Separately, the Dutch government said on Monday that it shares the United States’ security concerns regarding exports of advanced semiconductor manufacturing tools and is considering the latest U.S. rules.

The US Department of Commerce said the new regulations are aimed at slowing China’s development of advanced AI that could “change the future of warfare” and undermining the development of China’s own semiconductor ecosystem.

The agency said this was in line with Washington’s “small garden, high fence” policy of strategic restrictions, an approach that Chinese President Xi Jinping criticized last month.

Since the launch of ChatGPT raised global awareness of the power of AI, calls for further shutdowns of the semiconductor supply chain have been growing.

Thibault Denamiel, a fellow at the Center for Strategic and International Studies, told AFP that the move confirms “the trajectory of U.S. policy rather than a significant increase in regulatory efforts.”

“The additions become less important in light of the incoming Trump administration’s proposals,” he added, noting that the president-elect has vowed drastic action to trivialize these latest restrictions on chip technology.

with Agence France-Presse

Source: www.theguardian.com

Nvidia Continues Impressive Growth as AI Chip Leader Shows no Signs of Slowing Down

Nvidia, the AI chip maker that is the world's most valuable company and the driving force behind the artificial intelligence boom, delighted investors on Wednesday with new quarterly results.

The company's corporate value soared by $2.2 trillion this year to $3.6 trillion on the back of nearly doubling chip sales, but sales fell to $35.08 billion, compared to expectations of $33.15 billion. It was announced that it was worth $1,000,000. Profits more than doubled compared to the same period last year. Sales increased by 94% compared to the same period last year. The company predicted a 70% increase in revenue for the next quarter.

Analysts had expected NVIDIA to report earnings of $0.75 per share. The company reported $0.81. Nvidia stock fell about 5% in extended trading following the announcement, but quickly recovered its losses and maintained a similar price. The previous New York closing price was $145.89.

Nvidia CEO Jensen Huang said in a press release last week that he expects the computing power driving advances in generative AI to increase “a million times” over the next decade.

Huang said on the earnings call that the global adoption of Nvidia technology is causing a platform shift from coding to machine learning, and traditional data centers are being re-architected for machine learning to produce AI. said.

“Generative AI is not just a new software feature, it’s a new industry with AI factories producing digital intelligence, a new industrial revolution with the potential to create a multi-trillion dollar AI industry,” he said. .

“AI is transforming every industry, business, and country,” Huang added. “Creating an omniverse of synthetically generated data…the era of robotics is here.”

A surge in demand for Nvidia's Blackwell GPU chips appears to have eased fears that the company would be hit by a backlash in demand from tech giants, which are spending billions of dollars on AI processing and data centers. .

Nvidia stock has rebounded from its summer slump and is up 45% from its August lows. Semiconductor stocks are up nearly 200% this year, more than 1,100% over the past two years, and hit a record high after the election.

However, many of NVIDIA's chipmakers are struggling to compete with the advantages of AI, which is holding the industry back.

Before the results were released, Wedbush analyst Dan Ives said he expected more “drop-the-mic performance” from Nvidia, adding that “it's expected to generate more than $1 trillion in the next few years.” “This is the only game in the city where AI capital investment is planned.” Nvidia's GPUs will bring new oil and gold to this world. ”

The world's largest technology companies have increased their spending on AI by billions of dollars in recent quarters, with Nvidia positioned as a major beneficiary.

Nvidia is seen by many as a bellwether for the technology sector and artificial intelligence demand that helped propel Wall Street to multiple record highs this year.

Skip past newsletter promotions

But the escalation of the Russia-Ukraine war, the threat of global tariff hikes by the incoming Donald Trump administration and the possibility that the US Federal Reserve will not lower US interest rates are also spooking markets.

Other analysts echoed Ives' assessment that demand for Nvidia's new Blackwell chips could push Nvidia's sales and market capitalization to new heights. Charu Chanana, Saxo's chief investment strategist, said signs of “extraordinary demand” for new chips, including record sales forecasts and reports of sold-out inventory next year, are strong indicators that Nvidia's strong performance will continue. I wrote that.

But Chanana cautioned that “any signs of production delays or lack of demand could put pressure on the stock given valuation growth.”

Earlier this week, report Chipmakers say their latest graphics chips are having overheating problems with servers. www.theguardian.com

Review of Apple Watch Series 9 and Ultra 2: Improved chip speed and screen brightness

With brighter screens, new hands-free gestures, and faster speeds for the first time in years, Apple’s smartwatch has firmly established itself as the market leader.


The Apple Watch Series 9 will be available in a variety of sizes and materials, with prices starting at £399 (€449 / $399 / AU$649), a £20 reduction in the UK. It will launch alongside the Ultra 2, which costs £799 (€899 / $799 / AU$1,399), £50 cheaper than last year’s model.

Both watches look similar to their predecessors on the outside, with the Series 9 sporting a slimmer, pillow-shaped shape and available in 41mm or 45mm size options, while the 49mm Ultra 2 sports a chunkier look with a sturdy titanium shell and oversized buttons.

New this year are significantly brighter screens. The Series 9 is twice as bright as last year’s Series 8, with a maximum brightness of 2,000. LiceThat puts it on par with rival Samsung’s Galaxy Watch 6, and it has the same maximum brightness as the iPhone 15 Pro Max. A brighter screen makes it easier to read in direct sunlight, which is especially important for a watch.

The Ultra 2 goes even further, with a screen that can hit an impressive 3,000 nits at peak, which is 50% brighter than last year’s model. The super-bright screen is almost overkill unless you’re hiking through the desert, but you can unleash maximum brightness with the built-in torch mode to better light your way.

New, faster chips


Siri requests for apps, timers, and other simple actions are now much faster and more reliable, even when you’re out of range of your iPhone. Photo: Samuel Gibbs/The Guardian

The new S9 chip in both watches is the first to offer a significant speed boost since the S6 in 2020. The 30% performance increase isn’t dramatic for everyday activities, except for interactions with Siri, many of which are now handled on the watch. Setting timers, taking calls and other simple interactions are quicker without an iPhone or data connection. Voice input for messages and notes is also up to 25% more accurate, making it faster to send replies without a phone call.

Battery life for both models is around 36 hours for the Series 9 and 70 hours for the Ultra 2, enough for a full day and night, or nearly three days, on a single charge.

Double Tap is coming soon


The double pinch gesture is simple and easy to perform, making it especially useful for quick actions like silencing a timer when your other hand is occupied. Photo: Samuel Gibbs/The Guardian

Apple also added a new gesture for hands-free use of the Watch: “Double Tap,” which recognizes pinching your finger and thumb together twice, which is a simplified version of one of Apple’s existing gestures. AssistiveTouch accessibility features It’s for watches, but it’s faster and has become a standard part of every interface.

A double pinch will silence alarms and timers, start or end calls, and perform other basic actions. This feature only works when the screen is active, so the watch is facing you and can’t be accidentally activated. This feature requires the watchOS 10.1 update. It’s currently being tested in the public beta, but is working very well and should be fully rolled out by the end of October.

Sustainability

Apple doesn’t disclose the expected lifespan of the battery, but it is expected to last for more than 500 full charge cycles, retain at least 80% of its original capacity, and is replaceable. £95Repair costs Between £309 and £509 Varies by model.

They contain recycled aluminum, titanium, cobalt, copper, gold, plastic, rare earth elements, tin, and tungsten. Apple offers trade-ins and free recycling for its devices, and the report details the environmental impact of each watch.

price

The Series 9 comes in two sizes (41 and 45mm), a choice of materials, and a 4G option that requires an e-SIM-compatible phone plan add-on. Prices start from £399 ($399/AU$649), with the 4G model costing an extra £100 ($100/AU$160). The Ultra 2 costs £799 ($799/AU$1,399).

By comparison, the Apple Watch SE costs £259, the Samsung Galaxy Watch 6 costs £289, and the Google Pixel Watch 2 costs £269. £349The Garmin Venu 3 costs £449, while the Garmin Epix Pro costs £829.

verdict

Apple’s smartwatch didn’t need much to keep it in the top spot, but both the Series 9 and Ultra 2 look the same but with small but meaningful updates that are hard to beat.

The significantly brighter screen makes a big difference when you’re outside and want to see the time, alerts, and activity stats at a glance, and the faster chip means this watch will stay fast for years to come, making interactions with Siri super-fast, even when your phone isn’t nearby.

The upcoming double-tap gesture is also a handy addition, making tasks like checking train times a little easier when you’re rushing with your luggage.

If you want the best smartwatch for your iPhone, it’s hard to beat the Apple Watch, either model.


When the double-tap gesture is recognized, the watch vibrates and a notification pops up, here we are scrolling through widgets on the watch face. Photo: Samuel Gibbs/The Guardian

Source: www.theguardian.com

Sydney Researchers Lead the Way in Brain Chip Technology Ahead of Elon Musk’s Neuralink Neuroscience

BLaine computer interface technology is at the heart of movies like Ready Player One, The Matrix, and Avatar. But outside of the world of science fiction, BCIs are used on Earth to help paralyzed people communicate, to study dreams, and to control robots.

Billionaire entrepreneur Elon Musk announced in January that his neurotechnology company Neuralink had implanted the first computer chip in a human. In February, he announced that patients can now control a computer mouse with their thoughts.

Neuralink’s purpose is noble. It is about helping people who are unable to communicate or interact with their environment. But details are scant. The project quickly raised alarms about brain privacy, the risk of hacking, and other potential issues.



Dr Steve Kassem, senior research scientist at Neuroscience Research Australia, said the Neuralink news should be taken with a “large pinch of salt”. It’s not the first company to do neural implants, he says. In fact, Australia is a ‘hotspot’ for relevant neurological research.

Does the patient dream of electric sheep?

The University of Technology Sydney project, which has received millions of dollars in funding from the Department of Defense, is now in its third phase to demonstrate how soldiers can use brain signals to control robotic dogs.

“We succeeded [demonstrating] Handa can use his brain to issue commands that direct the dog to reach its destination completely hands-free…so the dog can use its hands for other purposes. ” he says.

Soldiers use assisted reality glasses with special graphene interfaces to issue brain signal commands to send the robot dog to different locations. Lin said he is working on making the technology multi-user, faster and able to control other vehicles such as drones.

Meanwhile, Sydney company Neurode has developed a headset to help people with ADHD by monitoring the brain and sending electronic pulses to help them cope with changes. Another his UTS team is working on it. dream machine, which aims to reconstruct dreams from brain signals. It uses artificial intelligence and brainwave data to generate images from your subconscious mind.

And then there are the implants.

good signal

Synchron started at the University of Melbourne and is now based in New York. it is, Mesh inserted into blood vessels in the brain This allows patients to use the Internet by transmitting signals that operate similar to Bluetooth. People can shop, send emails, and communicate online using technology that controls computers.



Synchron has implanted and monitored mesh in many patients, including one in Australia. Patient P4, who has motor neuron disease, had mesh implanted several years ago.

“I think he’s had over 200 sessions,” says Gil Lind, Sychron’s senior director of advanced technology. “He is still progressing well with his implant treatment and is working very closely with us.

“He was able to use the computer through the system…As the disease progressed, it became very difficult to use the physical buttons.

“This allows for online banking, communication with caregivers, [with] Someone I love. ”

Dr Christina Maher from the University of Sydney’s Brain and Mind Center said Synchron’s technology is “miles ahead” of Elon Musk’s, and is more sophisticated and safer as it does not require open brain surgery. Stated. The researchers have also published more than 25 papers, she said.

“As for Neuralink, we don’t know much about it.

“My understanding is that the top priority for them is to test the effectiveness and safety of surgical robots…so they are focusing more on the robotic side of things, and this is a commercial It makes sense from a perspective.”

Need for regulation

But amidst the hype and promise of neurotechnology, there are concerns about who will have access to the beneficial technologies and how they will be protected.

Maher says it’s important to balance the need for innovation with appropriate regulation while allowing access to those who really need it. She says the “gap between the haves and have-nots” is being discussed not just in Australia but around the world.

Skip past newsletter promotions


“As brain-computer interfaces become more common, people will be divided into those who can afford them and those who cannot,” she says.

Lind said Synchron is focused on those who have the most to gain, such as quadriplegic patients. “We want to expand it as much as possible. We hope to reach a bigger market and help more people in need,” he says.

A personal and pivotal moment for him, he says, was seeing the faces of the clinicians, team, and family of the first patient who received a successful implant.

At Neuralink, Kasem warns that there are always risks when technology is developed by a company that exists to make a profit. “A cell phone plan for the brain is not what we want,” he says.

“And what if this gets hacked? There’s always a risk when it’s not a closed system.”

But it’s more likely that Neuralink will use people’s data.

“Like every app on your phone or computer, Neuralink monitors everything it can. Everything it can,” Kasem says.

“It will be stored somewhere.”

Protect your brain data

Maher agrees that data is a big issue, saying the risk of hacking remains when devices are connected to the internet. She says much of the social media, biometrics, and other data is already out there, but her brain’s data is different.

“meanwhile [BCI companies] They are subject to the same data privacy laws…The difference in many people’s minds is that brain data is very private and it’s your personal thoughts.

“The big picture here is that once you start recording large amounts of brain data, there are absolutely megatons of data out there,” she says.

Despite privacy concerns, Kasem says interacting with the brain has exciting potential.

“We need to remember how powerful and important the brain is. All you are, all you have been, and all you will ever be is your brain and nothing else.” he says.

Quoting American physicist Emerson Pugh, he says the brain has trillions of neural connections that lead to “infinite opportunities.” hand. ”

Source: www.theguardian.com

‘Incredible Valor’: The Legacy of Grace Hopper in Nvidia’s Monumental $2 Trillion Chip Empire | Computing

I
In the demanding technical field of semiconductor manufacturing, hardcover book-sized processors stand out. Nvidia’s H-100. On Friday, the Santa Clara, Calif., company was valued at more than $2 trillion. The next step will likely be a chip named after U.S. Navy Rear Adm. “Amazing Grace” Hopper, who was instrumental in designing and implementing the programming language.


Nvidia supplies about 80% of the global market for chips used in AI applications. The company’s H-100 chips (the “H is for hopper”) are now so valuable that they have to be transported in armored vehicles, and demand is so great that some customers have to wait 6 months to receive it.

Hopper’s importance to Nvidia, and to AI computing more generally, was reinforced last summer when Nvidia founder and CEO Jensen Fan announced the next generation accelerated computing and generation AI chip, the GH200 Grace Hopper. It was emphasized when they named it a Super Chip.





Admiral Grace Hopper in 1985. Photo: Associated Press

Hopper was born in New York City in 1906, graduated from Vassar College in 1928 with degrees in mathematics and physics, and joined the Navy after the United States entered World War II following the attack on Pearl Harbor.

According to a biography from Yale University, Initially rejected by the Navy because of her age and small stature, she was commissioned and assigned to Harvard University’s Ship Bureau Computation Project, where she worked on the Mark I, America’s first electromechanical computer, calculating the rocket’s trajectory and reaction force, aircraft gun range table, and minesweeper calibration.

After the war, Hopper joined the Eckhart-Mauchly Computer Corporation (later Sperry Rand), where she pioneered the idea of automatic programming. In 1952, she developed the first compiler, a program that translated written instructions into computer code.

“What I was looking for when I started learning English [programming] was to bring in another whole group who could easily use computers. I kept asking for a more user-friendly language. Most of what we have learned from academics and computer science people has never been adapted to humans,” Hopper explained in a 1980 interview.

Skip past newsletter promotions

Hopper retired as a rear admiral at age 79, making her the oldest active duty officer in the U.S. military. The year before her death in 1992, she was awarded the National Medal of Technology by President George H.W. Bush. She was posthumously awarded the Presidential Medal of Freedom, the highest civilian honor, in 2016.

In a 1983 interview on “60 Minutes”, Hopper was asked if the computer revolution was over. Hopper replied: “No, we’re just getting started. I got a Model T.”

Source: www.theguardian.com