Artificial intelligence

Nvidia launches Spectrum-XGS to build global AI factories

American technology company Nvidia has unveiled Spectrum-XGS Ethernet, a new networking technology designed to connect multiple data centres into unified giga-scale AI factories.

With AI demand skyrocketing, single facilities are hitting limits in power and capacity, creating the need for infrastructure that can operate across cities, nations and continents.

Spectrum-XGS extends Nvidia’s Spectrum-X Ethernet platform, introducing what the company calls a ‘scale-across’ approach, alongside scale-up and scale-out models.

Integrating advanced congestion control, latency management, and telemetry nearly doubles the performance of the Nvidia Collective Communications Library, allowing geographically distributed data centres to function as one large AI cluster.

Early adopters like CoreWeave are preparing to link their facilities using the new system. According to Nvidia, the technology offers 1.6 times greater bandwidth density than traditional Ethernet and features Spectrum-X switches and ConnectX-8 SuperNICs, optimised for hyperscale AI operations.

The company argues that the approach will define the next phase of AI infrastructure, enabling super-factories to manage millions of GPUs while improving efficiency and lowering operational costs.

Nvidia CEO Jensen Huang described the development as part of the AI industrial revolution, highlighting that Spectrum-XGS can unify data centres into global networks that act as vast, giga-scale AI super-factories.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

NVIDIA eyes recovery in China after export deal ahead of Q2 report

NVIDIA is due to report its Q2 2026 financial results after the US market closes on 27 August, and analysts are expecting strong performance.

Consensus forecasts place revenue at around US $45.9 billion, up about 50 percent year-on-year, driven by ongoing demand for Blackwell GPUs, data centre expansion and redistribution of AI infrastructure investments globally.

Export changes are also pivotal. After entering a deal to resume H20 chip sales to China, despite revenue-sharing conditions, NVIDIA could reclaim as much as US$8 billion during Q2, mitigating past losses caused by restrictions.

Beyond geopolitical shifts, the Blackwell Ultra GPU is central to growth. Offering up to 50 times faster AI inference than earlier models, it is increasingly stocked by cloud providers and hyperscalers. Markets view this as a strategic advantage, fueling long-term AI momentum.

Risks remain. Gross margins may recover from prior pressure due to licensing charges, but margin expansion depends on supply and TAM realisation. China’s policy environment is also uncertain, making future guidance cautious for some analysts.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Dell expands AI innovation hub in Singapore to drive regional growth

Dell Technologies has launched a new Asia Pacific and Japan AI Innovation Hub in Singapore, strengthening its role in advancing AI across the region.

The hub extends the company’s Global Innovation Hub, which has already received more than US$50 million in investment since 2019. Its focus is on driving AI transformation, enablement and leadership, in line with Singapore’s National AI Strategy 2.0.

Instead of offering only infrastructure, the hub delivers end-to-end support, from strategy to deployment, helping enterprises bridge the gap between ambition and practical results. Research shows 62% of Singaporean businesses prefer such holistic partnerships.

Since 2024, the hub has developed about 50 AI prototypes and carried out more than 100 proof-of-concepts, workshops and demonstrations across areas such as generative and predictive AI.

The projects have already influenced multiple sectors. In energy, AI solutions are strengthening infrastructure resilience and enhancing customer engagement with digital humans and chatbots.

In telecommunications, AI is supporting agility and operational efficiency, while in education, cloud-based technologies are empowering research and innovation.

Dell’s AI Centre of Excellence Lab further supports these initiatives by testing solutions for AI PCs and edge computing in collaboration with academic and hardware partners.

A strong emphasis is also placed on skills development. By the end of 2025, the hub aims to train around 10,000 students and mid-career professionals in AI engineering, platform engineering and related fields.

Working with 10 local institutes, Dell is addressing the talent shortage reported by nearly half of Singaporean organisations. Events such as the Dell InnovateFest and the Dell Innovation Challenge provide platforms for students and partners to showcase ideas and create solutions for social good.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Scientists used AI model to anticipate fusion success ahead of experiment

Scientists at Lawrence Livermore National Laboratory employed a deep learning model to forecast the outcome of the December 2022 fusion ignition experiment at their National Ignition Facility (NIF).

The model, part of a cognitive simulation framework, predicted with approximately 74 per cent probability that the test would exceed the breakeven point, meaning it would yield more energy from fusion than the laser energy used.

The model combined over 150,000 high-fidelity simulations with real-world experimental data using Bayesian inference to enhance prediction accuracy. It ran on LLNL’s Sierra supercomputer and was embedded in their CogSim toolkit for physics-informed AI-driven simulation.

Researchers emphasise that this was not a lucky guess. The AI provided probabilistic forecasts complete with confidence intervals, guiding expectations and helping shape subsequent experimental designs.

Observers note that such an approach could dramatically reduce time and cost in future fusion experiments.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

OpenAI to open office in New Delhi

OpenAI has announced plans to open its first office in India later this year, selecting New Delhi as the location. India is now ChatGPT’s second-largest market after the US and continues to experience rapid growth in user activity.

Weekly active users of ChatGPT in India have increased over fourfold over the past year, with students making up the largest global user segment. CEO Sam Altman praised India’s talent pool and government support, stating the new office is key to building AI with and for India.

Union IT Minister Ashwini Vaishnaw welcomed the move, citing India’s AI mission and expanding digital infrastructure as a natural foundation for the partnership. OpenAI will also hold its first Education Summit in India later this month, aiming to further engage with students and educators nationwide.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Senior OpenAI executive Julia Villagra departs amid talent war

OpenAI’s chief people officer, Julia Villagra, has left the company, marking the latest leadership change at the AI pioneer. Villagra, who joined the San Francisco firm in early 2024 and was promoted in March, previously led its human resources operations.

Her responsibilities will temporarily be overseen by chief strategy officer Jason Kwon, while chief applications officer Fidji Simo will lead the search for her successor.

OpenAI said Villagra is stepping away to pursue her personal interest in art, music and storytelling as tools to help people understand the shift towards artificial general intelligence, a stage when machines surpass human performance in most forms of work.

The departure comes as OpenAI navigates a period of intense competition for AI expertise. Microsoft-backed OpenAI is valued at about $300 billion, with a potential share sale set to raise that figure to $500 billion.

The company faces growing rivalry from Meta, where Mark Zuckerberg has reportedly offered $100 million signing bonuses to attract OpenAI talent.

While OpenAI expands, public concerns over the impact of AI on employment continue. A Reuters/Ipsos poll found 71% of Americans fear AI could permanently displace too many workers, despite the unemployment rate standing at 4.2% in July.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Google claims Gemini uses less water and energy per text prompt

Google has published new estimates on the environmental footprint of Gemini, claiming a single text prompt uses about five drops of water and 0.24 watt-hours of electricity. The company says this equates to 0.03 grams of carbon dioxide emissions.

According to Google, efficiencies have reduced Gemini’s energy consumption and carbon footprint per text prompt by factors of 33 and 44 over the past year. Chief technologist Ben Gomes said the model now delivers higher-quality responses with a significantly lower footprint.

The company argued that these figures are significantly lower than those suggested in earlier research. However, Shaolei Ren, the author of one of the cited papers, said Google’s comparisons were misleading and incomplete.

Ren noted that Google compared its latest onsite-only water figures against his study’s highest total figures, creating the impression that Gemini was far more efficient. He also said Google omitted indirect water use, such as electricity-related consumption, from its estimates.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

Hong Kong deepfake scandal exposes gaps in privacy law

The discovery of hundreds of non-consensual deepfake images on a student’s laptop at the University of Hong Kong has reignited debate about privacy, technology, and accountability. The scandal echoes the 2008 Edison Chen photo leak, which exposed gaps in law and gender double standards.

Unlike stolen private images, today’s fabrications are AI-generated composites that can tarnish reputations with a single photo scraped from social media. The dismissal that such content is ‘not real’ fails to address the damage caused by its existence.

The legal system of Hong Kong struggles to keep pace with this shift. Its privacy ordinance, drafted in the 1990s, was not designed for machine-learning fabrications, while traditional harassment and defamation laws predate the advent of AI. Victims risk harm before distribution is even proven.

The city’s privacy watchdog has launched a criminal investigation, but questions remain over whether creation or possession of deepfakes is covered by existing statutes. Critics warn that overreach could suppress legitimate uses, yet inaction leaves space for abuse.

Observers argue that just as the snapshot camera spurred the development of modern privacy law, deepfakes must drive a new legal boundary to safeguard dignity. Without reform, victims may continue facing harm without recourse.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

Google launches Gemini AI for government

Google has introduced a new version of its Gemini AI platform tailored specifically for US government use, called Gemini for Government. The platform combines features such as image generation, enterprise search, and AI agent development, with compliance to standards like Sec4 and FedRAMP.

Gemini includes pre-built AI agents for research and idea generation, while also offering tools to create custom agents. US government customers will pay $0.50 per year for basic access, undercutting rivals OpenAI and Anthropic, who each launched $1 government-focused AI packages earlier this year.

Google emphasised security, privacy, and automation in its pitch, positioning the product as an all-in-one solution for public sector institutions. The launch follows the Trump administration’s AI Action Plan, which seeks to promote AI growth.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

South Korea unveils five-year AI blueprint for ‘super-innovation economy’

South Korea’s new administration has unveiled a five-year economic plan to build what it calls a ‘super-innovation economy’ by integrating AI across all sectors of society.

The strategy, led by President Lee Jae-myung, commits 100 trillion won (approximately US$71.5 billion) to position the country among the world’s top three AI powerhouses. Private firms will drive development, with government support for nationwide adoption.

Plans include a sovereign Korean-language AI model, humanoid robots for logistics and industry, and commercialising autonomous vehicles by 2027. Unmanned ships are targeted for completion by 2030, alongside widespread use of drones in firefighting and aviation.

AI will also be introduced into drug approvals, smart factories, welfare services, and tax administration, with AI-based tax consultations expected by 2026. Education initiatives and a national AI training data cluster will nurture talent and accelerate innovation.

Five domestic firms, including Naver Cloud, SK Telecom, and LG AI Research, will receive state support to build homegrown AI foundation models. Industry reports currently rank South Korea between sixth and 10th in global AI competitiveness.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!