NVIDIA has unveiled a factory-built modular system for AI data centres, aiming to cut deployment time from several months to as little as 14 days as demand for computing capacity surges.

Called DSX POD, the system integrates computing racks, networking, cabling, power distribution and cooling into factory-tested units shipped in transportable containers. Deliveries are scheduled to begin worldwide in early 2027.

“An AI factory creates value when it starts producing tokens, not when construction is complete,” NVIDIA founder and chief executive Jensen Huang said at the company’s GTC event in Berlin on Wednesday.

Tokens are units of text or other data processed and generated by AI models. For data centre operators, faster deployment means an earlier start to commercial operations and revenue generation.

Mr. Huang highlighted the financial implications of delays, saying a data centre expected to generate $40 billion in annual revenue would lose the equivalent of $10 billion if delayed by three months. The figures were illustrative estimates presented by him, not losses reported at a specific facility.

The DSX POD approach shifts much of the installation, integration and testing work from customer sites to factories, reducing the complexity of on-site deployment.

Built around NVIDIA’s Vera Rubin computing platform, the reference design includes seven liquid-cooled computing racks connected by 17 kilometres of Spectrum-X networking cables. The systems are fully cabled, integrated and tested before being shipped to customers.

“Once site utilities are ready, estimated on-site deployment goes from several months to days,” Mr. Huang said.

The 14-day timeline is conditional on site utilities being ready, meaning customers must still complete essential site preparation and infrastructure work before deployment.

NVIDIA said the factory-built approach would help operators bring computing capacity online faster, reduce the time equipment remains idle and limit capital tied up in infrastructure awaiting commissioning.

Union Minister for Environment, Forest and Climate Change, Bhupender Yadav, on Saturday interacted with representatives of start-ups and innovation companies working on air pollution mitigation, with a focus on developing practical, scalable and technology-driven solutions to address the air pollution challenge in Delhi-NCR.

Innovative approaches for low-cost mechanised road sweeping, monitoring road-dust resuspension, portable pothole repair and modular dust-containment systems were also discussed.

The Minister informed that the main focus of the interaction was on using artificial intelligence, sensors, remote sensing, satellite and drone-based technologies and data analytics to move towards real-time understanding of air pollution. He said that solutions presented by the innovators include hyperlocal air-quality monitoring, city-wide pollution mapping, hotspot identification and forecasting, AI-based source apportionment, real-time identification of pollution sources and predictive traffic-congestion management.

Technologies were displayed for 24×7 detection of stubble and open biomass burning, as well as automated identification of vehicles without valid PUC (Pollution Under Control) certificates and over-aged vehicles, which could further strengthen timely enforcement and intervention, he added.

Some of these companies are already working with different government bodies and more are expected to work with the government in different levels in the coming months, an official said.

A Bengaluru technology company says it is still waiting for $672,102 from an overseas customer for work completed more than 19 months ago, highlighting the financial risks smaller Indian technology exporters face when international clients fail to pay.

Thiyagu Ganesan, founder of Ticvic Technologies Pvt Ltd, has sought the Karnataka government’s intervention. His petition says the company completed the contracted work and the customer’s technical team tested it and confirmed that it was functioning.

The contract shared by Ticvic shows a second-phase project worth $672,102, with payments linked to milestones and customer acceptance. In an email dated May 29, 2024, a customer representative wrote: “I tested it and found that the NAT function is fine.”

Mr. Ganesan said Ticvic had received payment for an earlier project and, trusting the relationship, proceeded with the second phase without the expected upfront payment. He personally mortgaged land and pledged gold to keep all 47 employees paid, expecting to redeem the assets once the customer settled the dues. The company’s operations have since shrunk significantly, leaving only a few employees.

Ticvic describes itself as a network software engineering company with international customers and an overseas project dating back to 2018. It says it has approached China’s Ministry of Commerce, the CCPIT Mediation Center and the Indian Embassy in Beijing, but the dispute remains unresolved.

Recourse for Indian exporters when customers refuse to pay

Phil Fersht, CEO and chief analyst of HFS Research, said the remedies depend largely on the contract. Companies can pursue debt recovery, arbitration or litigation, enforce letters of credit or bank guarantees, or claim export-credit insurance where cover was arranged beforehand.

However, cross-border recovery can become prohibitively expensive for smaller firms. “Pursuing a $500,000 or $1 million claim through another country’s legal system can take considerable time and money,” he said. Once software is delivered and accepted, “the supplier has already surrendered much of its commercial leverage.”

“Winning a $672,000 overseas contract is not a $672,000 export success until you collect the $672,000,” Mr. Fersht said.

India has mechanisms to assist exporters. The Directorate General of Foreign Trade’s trade-dispute mechanism covers non-payment involving goods, services and technology, while the Export Credit Guarantee Corporation offers insurance for software and technology exports.

But prevention remains crucial. “Insurance and government support work much better when the exporter has conducted proper buyer due diligence, documented acceptance criteria, maintained the necessary records and structured the transaction correctly from the outset,” Mr. Fersht said.

Gaurav Vasu, founder and CEO of UnearthInsight, said non-payment was not a widespread problem across India’s IT services industry. Large companies such as TCS, Infosys and HCLTech have diversified customer bases and sophisticated legal and credit-risk systems. Smaller firms, however, can be devastated by a single default.

A chief delivery officer at a technology MSME said companies typically rely on legal counsel, collection agencies and negotiations rather than government intervention. His company is seeking to recover about $1 million, with discussions under way to accept equity in the customer’s business instead of pursuing litigation.

IBM is developing India-specific tools and application programming interface (API) connectors to integrate its sovereign AI software platform with local systems, as it seeks to help enterprises and governments retain greater control over their data, technology and AI operations.

At its Think 2026 event in Mumbai on Tuesday, Sandip Patel, Managing Director, IBM India and South Asia, described the approach as “digital swaraj”, arguing that sovereignty need not come at the cost of innovation.

“Sovereignty is not a choice between control and innovation. It is what gives enterprises the confidence to innovate at scale while remaining in control,” Mr. Patel said.

IBM’s Bharat Core comprises India-specific tools and API connectors built to work with its Sovereign Core software platform. Sovereign Core enables organisations to run cloud-like services on their own or partners’ infrastructure while retaining control over data, operations and AI workloads.

Bharat Core consists of around 20–25 tools and connectors designed to address Indian requirements, including integration with systems such as Aadhaar, Mr. Patel said during a press meet.

He cited three examples of IBM’s collaborations with Indian technology partners. IBM’s watsonx Orchestrate platform is available on Yotta’s Shakti Cloud, enabling organisations to deploy and govern AI agents while keeping data, operations and governance controls within India. IBM’s Sovereign Core has also been installed on NxtGen’s infrastructure to deliver secure, governed AI inference services in the country.

IBM’s collaboration with Indian AI company Sarvam combines Sovereign Core with Sarvam’s language, voice and reasoning capabilities, targeting government, public-sector and regulated applications.

Mr. Patel said sovereignty was gaining importance as enterprises and governments assessed how changing regulations and geopolitical developments could affect the infrastructure on which their operations depend.

“Sovereignty is not about building walls. It’s about controlling your keys,” said Hans Dekkers, General Manager, IBM Asia Pacific. Sovereign infrastructure, he said, involves equipping existing infrastructure with software and technology that allows organisations to retain control within India.

IBM’s approach centres on enabling organisations to run and manage AI on infrastructure they control, rather than requiring them to move all their workloads to public cloud platforms.

Separately, IBM on October 12 signed a memorandum of understanding with MeitY Startup Hub, an initiative of the Ministry of Electronics and Information Technology, to support India’s deep-tech startup ecosystem.

Under the agreement, startups will gain access to IBM’s enterprise technologies, technical guidance, mentorship, skilling programmes and opportunities to engage with enterprises and investors.

The collaboration will cover enterprise AI, hybrid cloud, quantum computing, sovereign AI and open-source ecosystem development, with a focus on startups from Tier-2 and Tier-3 cities. IBM and MeitY Startup Hub will also explore enterprise proof-of-concept projects to help startups validate and commercialise their technologies.

(The writer was in Mumbai at the invitation of the company)

Minister for Industries, IT and AI P.K. Kunhalikutty will formally launch MeitY IndiaAI Centre of Excellence in Artificial Intelligence Thiruvananthapuram, Keralam (CAIK), on Wednesday (September 30).

The centre is being established under the IndiaAI Mission of the Ministry of Electronics and Information Technology (MeitY), and the Department of Electronics and Information Technology. Digital University Kerala is the implementing agency and Innovation Incubator Advisory Pvt. Ltd. the industry partner. Transport Minister C.P. John will preside over the inaugural session.  

A major milestone accompanying the launch is the announcement of CAIK’s first cohort of 15 start-ups. 

India has not missed the artificial intelligence (AI) or quantum computing bus; rather, a perception largely shaped by overseas media suggests otherwise, Union Finance Minister Nirmala Sitharaman said on Saturday.

Speaking at an IIT Madras Alumni Association event, Sitharaman said the perception was partly driven by the absence of a significant number of AI companies listed on the National Stock Exchange or Bombay Stock Exchange. Several Indian AI startups, she noted, had already become unicorns or were close to achieving that status.

Indian startups were generally wary of listing because of the compliance requirements involved, she said, but pointed out that being listed also had significant advantages.

“Just not being in the stock market cannot be an adequate measure to see what’s happening in India. But mind you, that’s the world. We have to address or respond to these narratives too,” she said.

A stronger presence of AI companies on the bourses, she added, could signal to global investors that significant activity was taking place in India and attract more international funding.

Ms. Sitharaman called for greater investment in AI infrastructure, quantum computing and semiconductor manufacturing, which she described as the next major frontiers of growth. India, she said, needed to accelerate the development of hardware, skills and institutional capabilities rather than slow down.

“Despite the call for slowing down the speed with which innovators are coming up in the AI field, I’ll still think, for India, from where it is, we still need a lot more investment,” she said, particularly in hardware and in training teachers and institutions to continually upgrade and impart AI-related skills.

She also highlighted the growing need for AI among micro, small and medium enterprises across sectors to improve productivity and global competitiveness. The scale and timing of such investments, she said, was something academia and industry needed to determine together.

The next major wave of infrastructure investment, she said, would likely involve AI, semiconductors and quantum technologies, making collaboration between industry and academia critical to building capabilities.

Defence production, particularly in emerging technologies such as unmanned systems and drones, is another potential growth area, Ms. Sitharaman said.

Global defence suppliers themselves are struggling to keep pace with rapid technological change, she noted, creating opportunities for India to build capabilities and compete in expanding markets.

Finnish data operations company Aiven has announced the launch of its India operations, aiming to tap the country’s rapidly expanding artificial intelligence (AI), cloud computing and enterprise technology ecosystem.

The company, which provides open-source data infrastructure solutions for AI and cloud environments, will build “a local partner ecosystem to help Indian enterprises modernise their data platforms without being tied to a single technology provider or cloud environment.”

Aiven said its Bengaluru team, currently comprising 15 employees, will be expanded with a focus on sales, partnerships, solution engineering and other business functions. The company stressed that its India office would be a long-term investment rather than a marketing presence.

“India is increasingly where cloud, AI and enterprise innovation are being built at scale,” said Pranay Shree, one of the executives leading Aiven’s India expansion.

Ken Manuja, another Aiven executive, said “India has developer talent pool, digital businesses and cloud adoption momentum to become one of the key markets in the Asia-Pacific region for data transformation.”

Aiven’s clients include large firms across banking, fintech, cybersecurity, and telecom around the world.

The Bharat Sanchar Nigam Limited (BSNL) Deputy General Manager M. Satya Prasad on Saturday (September 19, 2026) asked all students to focus on artificial intelligence, robotics and quantum technologies to enhance their knowledge and build strong career opportunities.

Along with College chairman Muttamsetti Srinivasa Rao, he inaugurated freshers’ day celebrations in Avanthi’s St. Theressa Engineering College, Garividi off Vizianagaram district. Speaking on the occasion, he said that the students of all engineering branches can focus on certificate courses of new technologies from first year itself.

The Government Polytechnic College Principal, G. Satyanarayana, said that the students must focus on practical knowledge in chosen subjects to gain confidence and come out with new ideas.

Muttamsetti Srinivasa Rao said that the management had improved infrastructure and constructed new laboratories to make students to understand new technologies and do experiments frequently. Avanthi College Principal J.Bala Bhaskara Rao, Vice Principal B.Venkata Ramana, Administrative Officer G. Anil Kumar were present.

OpenAI will not go public in ​2026, Chief Executive Sam Altman said in remarks published on Saturday (September 12, 2026), citing safety ⁠concerns over artificial intelligence that render “unacceptable” even a 10% risk that AI could cause human extinction by decade’s end.

The comments come as more U.S. lawmakers urge new rules to govern AI systems after dire warnings from two ‌researchers from OpenAI rival Anthropic that rapidly progressing AI could lead to the extinction of the human race in the not-too-distant future.

“I actually think that, given everything ‌happening with safety, right now would be an ill-advised moment to go public, and we don’t ‌feel ⁠pressure on that,” Mr. Altman told business magazine Fortune.

Asked if there was a 10% ⁠chance of AI-driven extinction, Mr. Altman said he did not know how such an estimate could be made, but warned the risk was serious enough that AI companies and governments should act as though it could not be tolerated.

“Whether it’s 10 ​or eight or six, the point is, ‌we all have a tremendous amount of responsibility, and cannot let egos or incentives for profit or anything else get in the way,” Mr. Altman added.

“We need to act such that we are not taking any of those numbers of risk, and I believe we can.”

Instances of AI agents going rogue

Politicians, both Democrat and Republican, have responded with alarm and calls ​for more action after cases of AI agents going rogue to hack external systems and AI safety researchers quitting their companies out of concern at the technology ⁠risks.

In June, the New York Times said San Francisco-based OpenAI was considering whether to hold off until next year on a potentially trillion-dollar IPO.

At the time, shares in Elon Musk’s SpaceX IPO were ‌tumbling after a surge that sent that company’s valuation to $1.8 trillion.

Asked by Fortune if an IPO date of 2026 was off the table in favor of 2027, Mr. Altman said, “I would say not 2026.”

“We got a lot of stuff to do, like meeting this moment of what is going to be required for safety and alignment, and how the industry and governments can work together.”

Mr. Altman also suggested that OpenAI and other leading AI firms may be close to unveiling a pact to ‌slow AI development and work together to tackle safety risks, Fortune said.

Call for more deliberate approach

On Saturday (September 12), Anthropic ​CEO Dario Amodei urged AI companies to take a more deliberate approach to development.

“We must slow the pace at which we improve the capabilities of AI models,” Mr. ⁠Amodei wrote in an essay on X.

Mr. Altman later posted that he agreed with the sentiment.

“I agree ⁠with Dario that we need to pace the frontier,” Mr. Altman wrote. “This has been a primary topic of discussions we’ve had at OpenAI in recent weeks.”

Safety concerns have not slowed ‌Anthropic’s own IPO plans. It is expected to begin marketing its initial public offering in mid-October at the earliest and complete the listing days before the U.S. midterm elections in November, ​people familiar with the matter told Reuters this month.