Год: 2025

2025 was a breakout year for NVIDIA. The company became the first ever to hit the $5-trillion market cap, cementing its role as the backbone of the global AI infrastructure boom.

This year, NVIDIA expanded far beyond chips, backing more than 50 startups and major tech players with multi-billion-dollar commitments. Its investments span AI research, cloud infrastructure, autonomous systems, scientific computing, and even fusion energy.

Here’s a breakdown of where NVIDIA’s investment interests lie and which companies it aggressively backed this year.

OpenAI

OpenAI has been one of the headline partners for NVIDIA. In September 2025, the two companies announced an investment framework under which NVIDIA intends to invest up to $100 billion in OpenAI, starting in 2026. The capital infusion will help build and deploy at least 10 gigawatts of NVIDIA AI data centre systems.

The ChatGPT maker expects the first phase of data centres to come online in the second half of 2026.

Synopsys

In December 2025, NVIDIA announced a $2 billion equity investment in Synopsys, a semiconductor EDA (Electronic Design Automation) and IP leader, tied to a broad multi-year partnership.

This deal goes beyond typical venture financing. It integrates NVIDIA’s accelerated computing into design tools used across chip and system engineering for faster simulation and shorter design cycles.

Both companies said the partnership addresses rising workflow complexity, higher development costs, and pressure to shorten time-to-market across sectors, including semiconductors, aerospace, automotive, and industrial engineering.

Intel

NVIDIA made a major move by investing $5 billion in Intel through the purchase of common stock, giving it a significant equity stake in one of its longtime competitors.

Under the agreement, Intel will develop custom x86 CPUs for NVIDIA to use in its data centres, while also producing x86 system-on-chips that integrate NVIDIA RTX GPU chiplets for next-generation personal computers.

Nokia

NVIDIA announced a $1 billion investment in Nokia in October to support the development of AI-driven 5G and 6G networks, fuelling the chipmaker’s expansion into AI-focused telecom infrastructure. As part of the partnership, Nokia will integrate NVIDIA’s CUDA platform—a parallel computing platform and API model—into its radio access network software.

As 6G standards take shape, combining NVIDIA’s compute platforms with Nokia’s RAN expertise could accelerate the shift toward AI-native wireless networks.

Reflection AI

NVIDIA invested in Reflection AI as part of a huge $2 billion funding round that pushed the New York–based startup’s valuation to about $8 billion.

Reflection AI, founded by former DeepMind researchers, is building advanced AI systems that automate software development and other complex engineering tasks. Its models use large-context reasoning and agent-style workflows to solve complex autonomous coding challenges.

Mistral AI

Mistral AI, a leading European generative AI company, was part of NVIDIA’s investment portfolio in 2025. The French startup raised Series C funding of €1.7 billion (≈$1.99 billion), with NVIDIA participating in the round.

The companies also partnered to launch a new family of open source models.

Mistral’s models compete with major global LLM players and reflect NVIDIA’s strategy of supporting diverse AI model ecosystems across geographies.

Thinking Machines Lab

Thinking Machines Lab, an AI startup founded this year and led by former OpenAI CTO Mira Murati, raised a massive seed funding round of about $2 billion at a $12 billion valuation, with participation from heavyweights including NVIDIA, AMD, Cisco, and Jane Street.

The San Francisco-based startup is working on new kinds of AI systems that can handle many different tasks by combining advanced models with new system designs. The startup aims to build more reliable, safe, and accessible AI, unlike autonomous systems.

CoreWeave

NVIDIA increased its stake in CoreWeave, reaching about 24 million shares worth $3–4 billion by mid-year. This made CoreWeave one of NVIDIA’s biggest investments.

The stake grew from 17.9 million shares after the company’s March IPO to 24.28 million by August, as NVIDIA secured a key customer and GPU cloud provider as AI demand rises.

The investment sits alongside a $6.3 billion cloud capacity backstop deal signed in 2023 but revealed in September this year. It obligates NVIDIA to pay CoreWeave through 2032 for any unsold data centre capacity.

Perplexity AI

NVIDIA, among other ventures, backed Perplexity AI in its $100 million financing that skyrocketed the AI search startup’s valuation to $18 billion.​

This followed NVIDIA’s initial 2023 investment, with the chip giant joining heavyweights like SoftBank Vision Fund 2 and NEA to fuel Perplexity’s growth, including a massive $500 million infusion in late 2024.​

NVIDIA skipped the follow-on September $200 million raise at $20 billion valuation, in September this year, but ties remain strong through integrations in its sovereign AI projects for multilingual search capabilities.

Wayve

Wayve, a UK-based self-learning autonomous driving AI startup, signed a letter of intent with NVIDIA for a $500 million investment.

Wayve focuses on developing generalisable driving skills via machine learning. NVIDIA’s involvement supports AI compute inside vehicles and in cloud training.

Lila Sciences

Lila Sciences, which works on scientific superintelligence and AI-driven laboratory automation, raised $115 million in an extension round that saw participation from NVIDIA’s investment arm, NVentures, bringing its valuation to more than $1.3 billion.

This investment extends NVIDIA’s reach into AI-augmented scientific workflows, where massive datasets and complex models accelerate discovery in chemistry, materials science, and biology.

Commonwealth Fusion Systems

NVIDIA participated in an $863 million funding round for Commonwealth Fusion Systems, a nuclear fusion energy startup.

The investment supports the US startup in achieving commercial fusion power by the 2030s through its GPUs that offer high-performance computing for simulations.

Figure AI

NVIDIA is an investor in Figure AI, a US robotics company developing AI-powered humanoid robots to perform physical tasks in industrial, commercial, and domestic environments.

The company first raised major funding in a large $675 million round in February last year, with participation from NVIDIA, along with Microsoft, Jeff Bezos, and other technology investors.

This year, Figure AI continued to attract investment as it expanded its technology lineup, including its Helix AI system and BotQ manufacturing facilities, and reached a $39 billion valuation in a Series C funding round, with NVIDIA participating alongside other major partners.

The post The Top 13 Companies NVIDIA Bet Billions On in 2025 appeared first on Analytics India Magazine.

The United States and eight partner countries have launched the Pax Silica Initiative to build a secure and innovation-driven global silicon and AI supply chain.

The initiative brings together Japan, South Korea, Singapore, the Netherlands, the United Kingdom, Israel, the United Arab Emirates, and Australia. They plan to coordinate on critical minerals, semiconductors, AI infrastructure, energy, logistics, and manufacturing.

The initiative was launched at the inaugural Pax Silica Summit and outlined the coordination, why it is needed, and what actions will follow.

Pax Silica leaders said the goal is to reduce coercive dependencies, support trusted technology, protect critical materials, and enable partner nations to develop and deploy AI at scale. The US said countries have “affirmed a shared commitment to pursue projects to jointly address AI supply chain opportunities and vulnerabilities”.

The initiative responds to rising demand from partner countries for deeper economic and technology cooperation with the ỦS. Officials emphasised that AI is reshaping the world economy and will drive new demand for minerals, semiconductors, energy systems, and infrastructure.

Partner countries plan to work together on semiconductor design, fabrication and packaging, logistics, compute systems, minerals refining, and power generation. They will also explore joint ventures and co-investment opportunities, and seek to protect sensitive technologies from “undue access or control by countries of concern”.

The United States said its diplomats have been instructed to turn summit discussions into specific actions. According to the release, the US Under Secretary of State for economic growth, energy, and the environment, Jacob Helberg, has directed teams “to operationalise this summit’s discussions through identification of infrastructure projects and the coordination of economic security practices”.

The announcement noted that the countries will also “build trusted technology ecosystems, including ICT systems, fibre optic cables, data centres, foundational models and applications.”

Pax Silica takes its name from the Latin pax, meaning peace and stability, and silica, the base compound refined into silicon for computer chips. The initiative aims to unite countries that host major technology companies and investors to build a “secure, resilient, and innovation-driven ecosystem” across the global supply chain.

The post US Leads Pax Silica Initiative to Secure Global Silicon Supply Chain appeared first on Analytics India Magazine.

Indian enterprise AI is entering a new phase. For two years the narrative has been full of optimism, swelling venture capital, and a rush of pilots across sectors.

The country is still bullish on AI, but its enterprises are beginning to treat this technology like a true business asset. Leaders want results, not promises. Investors want resilience, not showmanship. Startups want customers who stick, not just early demos. CIOs want tools that work in their industry, not broad platforms that promise magic.

This shift is expected to test the industry’s confidence and define the next year for enterprise AI.

A sense of inevitability hangs in the air. A LinkedIn and Microsoft survey shows that 93% of Indian business leaders plan to deploy AI agents in the next 12 to 18 months. This is not tentative exploration, but a flood. Local experts see this acceleration as the result of long term preparation.

Kalyan Kolachala, MD at SAIGroup, a global enterprise AI leader, says, “India is very well prepared for growth in AI investments.” He points to a young talent base, cost advantages, policy support and major bets such as the $1 billion AI innovation centre in India. These factors indicate the country’s strides towards becoming an AI hub.

The demand is rising inside the enterprise and the supply side is maturing simultaneously.

A Filter, Not a Freeze

The course correction will begin inside India’s startup ecosystem. After an intense surge of AI funding through 2025, investors are preparing for a slower, more deliberate year. Sanchit Vir Gogia from Greyhound Research calls it a shift from high velocity to high conviction.

The tilt is already visible. Indian AI startups raised nearly $780 million in 2024, a 40% jump over the previous year, but early stage investments fell 37% as funds placed their trust in companies with proven traction. Greyhound’s data suggests that term sheets in 2026 will be tighter. Milestones will be stricter, and performance covenants will become common.

Greyhound sums it up: “This isn’t a freeze. It’s a filter… By 2026, the real signal will be resilience, not rhetoric.”

This discipline is being felt inside enterprises as well. The age of pilots with fuzzy metrics is ending. Many companies already treat AI as a governed asset with the same seriousness as ERP or CRM systems.

AI projects are now expected to come with a clear return on investment, transparent audit trails and continuous monitoring. CIOs are setting up internal policy boards that work with compliance and legal teams to manage error rates and risk. Greyhound captures the mood well: “Either AI earns its place on the revenue sheet, or it gets demoted to the backlog.”

Srinivas Reddy, senior vice-president and head of EPAM India, a global provider of software engineering and digital transformation services, agrees with this. According to him, the most important enterprise buzzword will be AI-native enterprise: not as a good to have, but as an operating reality. This will be the year organisations move from asking “how do we use AI?” to “how do we run the business with AI?”

AI-Native Enterprises

“Now that AI has entered the revenue sheet, it will slowly make strides towards an ‘innovation spend’ in 2026,” Reddy says. “Organisations will look at it as a core business capability, just like cloud, security or data platforms. Boards will measure AI by impact: revenue contribution, time-to-market reduction, engineering productivity and operational resilience.”

Reddy says that EPAM expects four focus areas to dominate conversation and investment: AI-native engineering, agentic workflows, enterprise-grade GenAI copilots and responsible AI at scale.

The idea of AI as a playground for experimentation is fading. In 2026, these systems must tie themselves to revenue, cost savings, or compliance, or step out of the way. Sustainability pressures are also reshaping decisions. Energy firms now plan their training workloads around renewable power availability. Boards are tracking model right sizing as a serious metric.

Indian IT firms have also called out their increasing AI ROI from investments. More clear revenue might finally be visible across all firms.

What is most striking is that enterprises are no longer buying AI as a general purpose tool. They want vertical specificity. Across banking, manufacturing, healthcare, retail and logistics, the demand is rising for tools that come pre aligned with sector knowledge.

A bank does not want to spend months training a model on KYC. A hospital does not want AI that cannot show how it arrived at a diagnosis. A factory wants quicker throughput, not a platform that needs a year of integration. This is why vertical AI is expected to dominate the coming year.

Reports from the last cycle already show sector specific gains in areas like fraud detection, diagnostic imaging, and yield optimisation. Greyhound notes that enterprises have grown wary of platforms that offer everything and prefer solutions that solve measurable problems in their industry.

The next wave of winning startups will be the ones with deep domain insight, not the ones with the widest pitch decks.

“Vertical AI will be a major focus area. Sectors like legal, healthcare, insurance and security are full of delays, manual work and large backlogs,” Ujwal Sutaria, founder and general partner at TDV Partners, an early stage venture firm, said. “India has millions of pending court cases and an overburdened healthcare system. AI can help speed up processes, reduce errors and improve access.”

Sutaria said that startups building in India for the world will also become more prominent. “Whether it is in AI, financial services or consumer brands, Indian companies have a strong opportunity to go global. Just like India became the IT services hub for the world, it has the potential to become a major AI hub,” he said.

Sutaria said that the funding climate for 2026 carries a sense of disciplined optimism. After the turbulence of 2023 and 2024, India’s startup world is entering a rebalancing year where money moves toward models that work, profits matter more than breakneck expansion, and founders with a record of building steady companies get the advantage.

These pressures vary by industry, but all of them point in the same direction. Banks now insist on usage based contracts and full model traceability. Hospitals demand interoperability, explainability and clear lines of liability. Manufacturers want payback in two or three years, not distant promises.

Energy companies tie AI operations to renewable goals. Retail wants systems that work instantly with existing supply chains. Many old style software contracts are being reopened and renegotiated because the expected AI gains have not materialised. The theatre of transformation is over. An AI product that cannot move a central KPI will have no place in the contract.

What About Sovereignty?

This maturity of enterprise AI is also redefining India’s approach to AI sovereignty. The debate is not about isolation, but pragmatism.

Vikas Singh, chief growth officer at Turinton, a business consulting and services firm, captures this tension with clarity. “Sovereignty in AI is a legitimate concern, but we already have a proven playbook right here in India. Look at IT services. That industry didn’t win by reinventing the wheel or building inferior alternatives. It won by leveraging global best practices, assembling them smartly, and solving complex enterprise problems at scale.”

Singh adds that the real question is not whether a tool is Indian, but whether it solves the business problem better. Differentiation, he says, should come from architecture and domain expertise. He calls for pragmatism rather than protectionism. “We don’t need to reinvent foundational AI models. We need to build platforms that solve enterprise complexity better than everyone else.”

Jaspreet Bindra, co-founder of AI and Beyond, a firm providing AI and tech literacy to organisations, shares a similar view, while adding a warning. “India’s push for AI sovereignty is both timely and strategic, but it must not come at the expense of quality or innovation.”

He says sovereign models are making progress in linguistics and culturally relevant datasets, but parity is still some distance away. And that is fine because every ecosystem matures gradually. The danger, he says, lies in forcing enterprises to use local tools that are not ready. “The real win for India will be when sovereign AI and world class performance become the same thing.”

Indian enterprises will adopt AI faster than many regions because these models will be fit for purpose and cost efficient. The only condition is that the ecosystem must focus on depth, not just announcements.

India enters 2026 with clear advantages. Nearly all major enterprises have already begun their AI journeys. The government is pushing supportive policies. Talent is abundant. Data centres are expanding.

Global and local players are investing at record scale. The coming year will not be defined by how many pilots begin. It will be defined by how many will scale, survive scrutiny and prove their value.

Indian enterprise AI is growing up. The optimism is still around, but it now rides on discipline. That combination is what will drive the country’s next leap.

The post 2026 Could be India’s Year in AI, But Only the Resilient Will Survive appeared first on Analytics India Magazine.

The Walt Disney Company (Disney) has accused Google of large-scale copyright infringement, according to multiple media reports citing a copy of a cease-and-desist letter the company sent on December 11.

The letter, reviewed by outlets including Variety and Axios, alleges that Google copied a substantial portion of Disney’s copyrighted catalogue without permission to train its generative AI models and then used those models to reproduce and distribute Disney-owned characters, images, and other creative assets.

According to the reports, Disney’s counsel at Jenner & Block wrote that Google is infringing the company’s copyrights “on a massive scale” by ingesting Disney content into its AI training pipelines and by outputting derivative works through services such as Gemini, Imagen, and Veo.
The letter, as reported by Variety, describes Google’s AI products as functioning like a “virtual vending machine” that can generate unauthorised images and renderings of characters from franchises including Frozen, Moana, The Lion King, Marvel, Pixar, and Star Wars.

Some outputs, Disney argued, appeared with Google’s branding, creating the false impression that the reproductions were licensed.

Disney has demanded that Google immediately stop copying or generating content derived from its intellectual property and restrict the availability of such outputs on YouTube, Shorts, and other Google platforms.

The company also noted in the letter that it had raised concerns with Google over several months but saw no meaningful corrective action.

The confrontation landed during the same week Disney announced a billion-dollar licensing and investment deal with OpenAI that will allow its characters to appear in Sora and other OpenAI media tools under formal commercial terms.

This will allow Sora, OpenAI’s generative video platform, to create short, user-prompted social videos featuring more than 200 characters from Disney, Marvel, Pixar and Star Wars.

The deal also includes a $1 billion equity investment in OpenAI, along with warrants to purchase additional equity.

The transaction remains subject to definitive agreements, corporate and board approvals, and other closing conditions. Sora and ChatGPT Images are expected to begin generating Disney-licensed content in early 2026.

The post Soon After $1 Bn OpenAI Deal, Disney Accuses Google of AI Copyright Infringement appeared first on Analytics India Magazine.

The Telangana government on Thursday announced global majors, including Costco and Stolt-Nielsen, are reportedly establishing their global capability centres (GCCs) in Hyderabad, further strengthening the city’s position as a premier technology hub. The announcement comes soon after the Telangana Rising Global Summit at Bharat Future City.

American retail giant Costco, with annual revenues of $270 billion, will set up a GCC that is expected to scale to over 500 employees. The Hyderabad centre will act as the company’s global technology backbone, supporting digital platforms, data and AI initiatives, supply chain systems and enterprise operations.

UK-based bulk logistics leader Stolt-Nielsen will launch a digital innovation centre in the city. The facility will become its global hub for product development, DevOps and next-generation digital capabilities spanning software engineering, data and analytics, automation, and operational systems.

Meanwhile, US financial services multinational Western Union confirmed the establishment of its second India GCC in Hyderabad, focused on platform engineering, advanced digital transformation and strategic technology initiatives.

The centre, developed in partnership with HCLTech under a Build-Operate-Transfer (BOT) model, is expected to employ over 400 professionals.

Rajeev Mago, head of India technology centres at Western Union, said the upcoming facility will play a key role in accelerating the company’s long-term digital expansion.

Senior leaders from all three companies cited Hyderabad’s world-class infrastructure and deep talent pool as significant factors influencing their decision to invest in the city.

Meanwhile, Google and the Telangana government have launched the Google for Startups Hub at T-Hub in Hyderabad, which includes a dedicated space to support the growing startup ecosystem.

Google will engage with regional startups from Telangana through a dedicated Hub, providing free year-long coworking spaces for selected AI-focused startups and access to curated venture investors.

The post Costco, Western Union & Stolt-Nielsen to Set Up GCCs in Hyderabad appeared first on Analytics India Magazine.

Meta has appointed Aman Jain as its new head of public policy in India, the company announced on December 11, 2025. Jain will join early next year and report to Simon Milner, vice president of policy for Asia Pacific (APAC). He will also be part of Meta’s India leadership team.

Jain brings over two decades of experience in public policy and business strategy. He has previously worked with Amazon, Google, the Government of India, and international bodies.

At Google India, he served in senior roles, including country head for Government Affairs & Public Policy. Most recently, he was director of public policy at Amazon, where he led policy strategy across marketplace, operations, competition, and technology.

Confirming the appointment, Milner said, “India is a strategic market for Meta. As the country’s digital economy accelerates across areas such as AI, emerging tech and the creator economy, Meta aims to help build a more inclusive, trusted, and future-ready internet ecosystem for India.”

Jain will lead Meta’s policy strategy and engagements in one of its largest markets. Milner said Jain’s background in public policy and technology would support Meta’s efforts to work more effectively with regulators and industry groups on shaping a conducive policy framework, and that he is expected to strengthen the APAC policy leadership team.

The post Meta Appoints Aman Jain to Oversee Public Policy in India appeared first on Analytics India Magazine.

Drone tech startup ideaForge Technology and the Centre for Development of Advanced Computing (C-DAC) have signed a Memorandum of Understanding (MoU) to integrate drones into India’s emergency response system.

The move aims to cut response time by linking ideaForge’s FLYGHT drone network with C-DAC’s Emergency Response Support System (ERSS), also known as Dial 112. The firms will also conduct research on UAVs (unmanned aerial vehicles), semiconductors, and data technologies.

The MoU also sets a plan to evaluate the C-DAC’s VEGA Processor chips family for UAV use, explore system-on-chip designs and study autonomous drone swarms powered by AI. Both groups will run joint training and research programmes to support India’s deep tech push.

The partnership will use the FLYGHT CLOUD platform—a cloud-based drone data management and analytics solution—as a rapid first response tool. It aims to address how agencies can deploy drones to provide real-time information to police, fire, and medical teams. The MoU outlines how both organisations will collaborate, what technologies will be explored and how integration will function in live operations.

Sachin Pukale, AGM, product management at ideaForge, said in a statement, “Integrating FLYGHT with ERSS will allow Automated Aerial Dispatch of drones to reach incidents within minutes and provide critical situational awareness intelligence to responders on the ground.” He added that the system supports data-driven decisions through an open architecture that can link with third-party tools.

ideaForge said emergency teams will gain quicker situational awareness when drones reach an incident before ground units. The company stated that FLYGHT allows agencies to use drone services without owning hardware or training specialised staff, as well as enabling state governments to adopt drone support at scale.

C-DAC said the collaboration will support the secure handling of drone data for public safety operations. It noted, “Our collaboration with ideaForge opens new opportunities to harness drone-generated data for faster decision-making, especially in emergency response scenarios.”

ideaForge and C-DAC said the partnership aligns with national goals for secure digital infrastructure and indigenous technology. They aim to support the Aatmanirbhar Bharat initiative by providing UAV platforms that meet safety and data governance requirements.

The post C-DAC, ideaForge Sign MoU to Use Drones for Faster Emergency Response appeared first on Analytics India Magazine.

At the Microsoft AI Tour held in Mumbai on December 12, Microsoft CEO Satya Nadella announced MahaCrimeOS AI, a platform that offers AI tools to fight cybercrime.

The platform, powered by Microsoft’s cloud platform Azure, is developed by CyberEye, a network security company and an independent software vendor of Microsoft, along with Maharashtra government’s special purpose vehicle MARVEL, and the Microsoft India Development Centre.

MahaCrimeOS AI is currently live in 23 Nagpur police stations, and Maharashtra Chief Minister Devendra Fadnavis proposed its future expansion to all 1,100 police stations across the state.

The platform supports officers by handling routine investigative tasks—creating cases instantly, extracting information across languages, and providing contextual legal guidance.

The platform brings together multiple AI assistants, automated workflows, and secure cloud infrastructure.

It also enables built-in access to India’s criminal laws using retrieval-augmented generation (RAG), a technology which pulls relevant statutes and precedents from approved sources to ground AI responses in accurate legal material.

It helps investigators to link related cases, analyse digital evidence, and respond to emerging threats more quickly and precisely.

“Our collaboration with Microsoft began with solving complex cybercrime challenges, but its potential is far greater,” said Fadnavis.

“AI today touches every sphere of human activity, from healthcare and agriculture to industry and governance, and we intend to harness this power responsibly to create a more effective, citizen-centric state.”

Ram Ganesh, CEO, CyberEye, said, “Our collaboration with Microsoft and MARVEL has enabled us to empower cutting-edge officers even in remote parts of the state to solve complex cybercrime investigations with ease and reduced workloads.”

Using the platform, Microsoft stated that FIR creation dropped to 15 minutes due to automated data extraction. Tasks that previously took 2–3 months are now completed in about a week.

Besides, investigators who could earlier manage one case per month now handle 7–8 cases.

In the last few days, Nadella was present in India, where the company hosted an AI tour across Delhi, Bengaluru and Mumbai.

Notably, the company announced a $17.5 billion investment in India over four years (2026–2029) to expand cloud and AI infrastructure, skilling programmes and ongoing operations.

The company also said it will expand partnerships with Cognizant, Infosys, Tata Consultancy Services, and Wipro, positioning the four Indian IT majors as “frontier firms” in the global adoption of agentic AI.

Each company will deploy more than 50,000 Microsoft Copilot licenses, collectively exceeding 200,000 seats, the company said.

The post All Maharashtra Police Stations to Get Microsoft-Powered AI Cybercrime Tool appeared first on Analytics India Magazine.