RTX Spark, unveiled on Monday by Jensen Huang, chief executive during a keynote in Taipei ahead of the Computex technology conference, brings about a shift in how computers are used, moving away from traditional software-based workflows towards systems that can carry out tasks through AI agents.

“The PC is being reinvented,” Huang said. “For forty years, you launched apps. Click. Type. With RTX Spark and Microsoft Windows, you ask, and the PC does the work.”

RTX Spark is designed as a superchip built for what Nvidia describes as the “era of personal AI agents”. It combines a Blackwell-based GPU with a Grace CPU, delivering up to 1 petaflop of AI performance and 128GB of unified memory. 

Nvidia says this setup is intended to support complex AI tasks running directly on the device rather than in the cloud.

The company explained that the chip will allow users to run large language models locally, including systems with up to 120 billion parameters, while also handling demanding creative and gaming workloads. These include editing high-resolution video, generating AI video content, and running advanced 3D rendering tools.

Nvidia said RTX Spark systems will support Windows PCs built for what it calls “personal agents”, software that can carry out tasks across applications. The company is working with Microsoft to integrate the technology into Windows, including new security features designed to control how AI agents operate on a device.

Microsoft chairman and chief executive Satya Nadella said the collaboration aims to expand access to advanced computing tools. “Our goal is to deliver unmetered intelligence to every home and every desk with Windows,” he said.

The companies noted that the new Windows platform will include tools that allow users to manage what AI agents can access, how data is handled, and when information is processed locally instead of being sent to the cloud.

RTX Spark also targets creators and developers as Nvidia said the chip can support 90GB 3D scene rendering, 12K video editing, and AI-assisted design work. Users will be able to run high-end gaming titles at 1440p resolution with frame rates above 100 frames per second.

Adobe is among the companies adapting its software for the new system. It is reworking Photoshop and Premiere to take advantage of the hardware, with expected performance gains in AI tools such as generative editing and video expansion features.

Shantanu Narayen, Adobe’s chair and chief executive, said the changes would speed up creative work. “The best creative work in the world happens in Adobe tools from Adobe Firefly to Photoshop and Premiere, and the expansion of our partnership with NVIDIA and Microsoft will make those experiences faster and more powerful than ever,” he said.

Other software and gaming companies are also involved, including Blackmagic Design, Blender, ComfyUI, OTOY, and Xbox, all of which said they are preparing support for the new platform.

Hardware makers are preparing devices around the chip. Nvidia said laptops and compact desktops will be produced by companies including ASUS, Dell, HP, Lenovo, Microsoft Surface and MSI, with Acer and GIGABYTE also expected to join later. The first devices are scheduled for release in the autumn.

RTX Spark systems are expected to come in slim laptop designs and compact desktops aimed at both professionals and consumers. Nvidia said laptops will feature lightweight builds, OLED displays and all-day battery life.

With the launch, Nvidia is going beyond its traditional graphics chip business into full PC system design. Analysts say the move places the company in closer competition with Intel, AMD and Apple in the personal computing market.

With this, Dell could add more than $80 billion to its market value.

The company reported first-quarter revenue of $43.8 billion for its 2027 fiscal year, far ahead of analyst expectations of about $35 billion.

Earnings also came in stronger than expected, with adjusted earnings per share reaching $4.86 compared with forecasts of roughly $2.94. Net income climbed 256% year-on-year to $3.44 billion.

The strongest growth came from Dell’s Infrastructure Solutions Group, the division responsible for servers and data-centre systems. Revenue in that business jumped 181% to $29 billion as companies continued spending heavily on AI infrastructure.

Dell noted that AI server sales alone reached $16.1 billion during the quarter, up 757% from a year earlier.

The company also booked $24.4 billion in new AI server orders, pushing its backlog to $51.3 billion. That means Dell still has tens of billions of dollars’ worth of systems waiting to be delivered over the coming quarters.

Investors focused heavily on the growing backlog because it gives Dell unusual visibility into future demand at a time when many technology companies still find it difficult to predict how long the AI spending wave will last.

Following the results, Dell raised its full-year revenue forecast to between $165 billion and $169 billion, up from earlier guidance of $138 billion to $142 billion.

The company now expects AI server revenue to hit roughly $60 billion this year, compared with its previous estimate of $50 billion.

Adjusted earnings per share guidance also increased to $17.90 from $12.90.

Dell’s recent growth has been tied to Nvidia, whose graphics processors power most of the company’s AI systems. The results came only days after Nvidia itself reported record data-centre revenue of $75.2 billion, up 92% from a year earlier.

Together, the numbers from both companies point to aggressive spending on AI infrastructure by major technology firms and cloud providers.

Dell has benefited from orders linked to companies including Alphabet, Amazon and CoreWeave, as well as large AI data-centre projects in Europe.

The company has also expanded its partnership with Nvidia through Project Helix, an initiative designed to help businesses build and deploy AI systems more quickly.

The latest earnings added to signs that demand for enterprise AI infrastructure remains strong across the industry. Lenovo recently reported strong AI server growth, while Super Micro Computer continues expanding manufacturing for GPU-based servers.

Technology companies are expected to spend hundreds of billions of dollars on AI data centres this year as competition around AI services intensifies.

Before the recent shares surge, Dell spent years being seen mainly as a PC and storage company. Now, investors view it as one of the major suppliers benefiting from the global vision to build AI infrastructure.

The Japanese investment company had originally aimed to secure a $10 billion margin loan backed by its OpenAI holdings.

However, discussions with banks and other potential lenders have recently shifted towards a smaller deal that could fall to about $6 billion, according to people familiar with the talks.

The loan is still under discussion and the final size could still change.

Lenders reportedly became cautious because OpenAI is privately owned, making it harder to determine a stable market value for the company.

Although OpenAI was recently valued at around $852 billion in a funding round earlier this year, creditors are wary about using unlisted shares as collateral for such a large borrowing arrangement.

A margin loan allows investors to borrow money against the value of assets they already own. In this case, SoftBank planned to use its OpenAI stake to secure the financing.

The proposed loan would run for two years, with an option to extend it by another year. Reports earlier this year also said the borrowing could carry an interest rate tied to SOFR plus 425 basis points, pushing costs close to 8%.

That is significantly higher than standard corporate lending rates and reflects the risks lenders see in the structure.

SoftBank has increased its financial exposure to OpenAI over the past two years. The company first invested in the ChatGPT maker in September 2024 and later expanded the partnership through Stargate, a large artificial intelligence infrastructure project launched in the United States in January 2025.

In March this year, SoftBank also secured a separate $40 billion bridge loan backed by major banks including JPMorgan and Goldman Sachs.

The company said the funding would support OpenAI investments and broader corporate operations.

Analysts estimate SoftBank’s total investment commitment to OpenAI could eventually reach about $64.6 billion, giving the group roughly a 13% in the company.

At the same time, some analysts believe SoftBank faces a financing gap of around $32 billion over the next two years.

To raise cash, the company has already sold several major assets. In 2025, SoftBank exited its Nvidia position for about $5.8 billion and also sold T-Mobile shares valued at roughly $12.7 billion.

Credit rating agency S&P recently revised SoftBank’s outlook to negative while keeping its BB+ rating, pointing to the company’s debt exposure and aggressive borrowing strategy.

Neither SoftBank nor OpenAI immediately responded to requests for comment following the latest reports on the loan discussions.

The agreement will see Meta use tens of millions of Graviton processing cores, according to Amazon Web Services executive Nafea Bshara, who said the contract would run for several years and be worth billions of dollars.

Demand for AI infrastructure is spreading beyond graphics processors made by firms such as Nvidia, while GPUs are essential in training AI models. Companies now need large volumes of central processing units to run trained systems, manage workloads and support AI agents.

Meta said the deal is part of its strategy to avoid relying on one supplier or one type of chip.

“As we scale the infrastructure behind Meta’s AI ambitions, diversifying our compute sources is a strategic imperative,” Santosh Janardhan, head of infrastructure at Meta, said in a statement.

Amazon said Meta chose its latest Graviton5 processor because of its price and performance. The chip is Amazon’s fifth in-house CPU generation and is produced by Taiwan Semiconductor Manufacturing Co.

“We pass that savings on to the customers,” Bshara told Reuters.

He added that most of the chip capacity for Meta would be based in the United States.

The partnership builds on an existing relationship between both companies that dates back several years. Earlier work had focused mainly on cloud services, Amazon’s Bedrock platform and GPU rentals.

For Meta, the latest agreement adds to its list of chip partnerships. The company has already signed major supply deals with Nvidia and AMD, while also working with Arm Holdings.

Amazon, meanwhile, is going deeper into AI infrastructure with both its own silicon and outside partnerships. Earlier this week, it announced another $5 billion investment in Anthropic, which will also use tens of millions of AWS Graviton cores.

The iPhone maker said on Monday that Ternus will take over on September 1, while Cook will become executive chairman.

The leadership change comes as Apple strengthens its focus on artificial intelligence, responding to competition from companies including Nvidia, Meta and Google.

Ternus joined Apple in 2001 and currently serves as senior vice-president of hardware engineering. He has worked on several of the company’s biggest products, including the Mac, iPad and AirPods.

He is also seen as an important figure in improving Mac sales in recent years, helping the product regain momentum against personal computer competitors.

Although he has kept a lower public profile than some Apple executives, the company has recently given him a more visible role.

Last year, Ternus presented the iPhone Air, a major redesign of Apple’s flagship device and one of the biggest changes to the product line in years.

At 50, he takes over at the same age Cook did when he succeeded Apple co-founder Steve Jobs in 2011.

Cook leaves the chief executive role after overseeing one of the most successful periods in Apple’s history. Since taking charge in August 2011, he has helped increase the company’s market value by about $3.6 trillion.

He was widely credited with expanding Apple’s global supply chain, especially through manufacturing partnerships in China, while also growing the company’s services and hardware businesses.

Cook also became the first Fortune 500 chief executive to publicly come out as gay in 2014 and often spoke on issues including workplace diversity and environmental policy.

Apple said Cook will remain involved in dealing with policymakers as executive chairman.

Ternus now inherits a company under pressure to show stronger progress in artificial intelligence.

Although Apple introduced Siri in 2011, it has struggled to match the pace of newer AI-focused companies.

Tech giants such as OpenAI and Anthropic have attracted millions of users with new chatbot products, while Nvidia has become the world’s most valuable listed company on the back of demand for AI chips.

In January, Apple reached an agreement with Google to use Gemini technology to improve Siri.

Ternus will also face competition in new devices. Meta Platforms has found success with smart glasses, while Apple’s Vision Pro headset has faced questions over its high price.

Alongside appointing John Ternus as CEO, Apple said Johny Srouji has been named chief hardware officer. He will continue leading the company’s custom chip and sensor teams.

The hardware engineering group previously led by Ternus will now be overseen by Tom Merieb.

The company confirmed on Thursday that it will not proceed with the British phase of the project for now, saying work will resume only when conditions support long-term investment.

Stargate UK, developed with Nvidia and British developer Nscale, was announced in September 2025 as part of a plan to expand global data centre capacity.

The project was expected to deploy up to 31,000 AI chips and strengthen the country’s ability to run its own artificial intelligence systems.

That capacity, usually called sovereign compute, allows a country to manage sensitive data and AI workloads locally instead of relying on overseas providers.

OpenAI said in a statement: “We see huge potential for the UK’s AI future. AI compute is foundational to that goal, we continue to explore Stargate UK and will move forward when the right conditions such as regulation and the cost of energy enable long-term infrastructure investment.”

The decision is a setback for the UK government as Prime Minister Keir Starmer has made artificial intelligence central to his economic plans and wants Britain to attract more global tech investment.

Officials insist talks are still ongoing. A spokesperson said the government is “continuing to work with OpenAI and other leading AI companies to strengthen UK compute capacity”.

At the same time, they pointed to more than £100 billion in private investment that has flowed into the UK’s AI sector since 2024.

The cost of energy is also a big issue. Britain has some of the highest electricity prices in Europe, and large data centres require vast amounts of power to run and cool advanced chips. Regulation is another concern, especially those around data use and copyright.

OpenAI has been expanding its data centre footprint in other regions. Its Stargate programme includes projects in the United States, Norway and the United Arab Emirates. The first major campus is already underway in Texas.

The pause in the UK also comes as the company strengthens its focus. It has scaled back some side efforts and is concentrating more on core services like ChatGPT.

Competition is increasing, with companies such as Anthropic and Google pushing ahead with their own systems.

Despite the Stargate project delay, OpenAI says it will continue discussions with the UK government, including plans to support public services with its technology.

For now, the project is on hold, with no timeline for when work might begin.

The company said on Tuesday it is projecting an operating profit of 57.2 trillion won ($38.2 billion) for the January to March period.

This is far ahead of expectations and more than eight times higher than the 6.69 trillion won it reported a year earlier. It also exceeds the company’s total profit for all of last year.

This would be Samsung’s strongest quarterly result on record. Its previous high stood at 20 trillion won, reached in the final quarter of 2025.

Demand from data centres has pushed prices higher. Companies building AI systems buy large volumes of memory chips, stretching supply. As a result, prices for DRAM chips rose sharply in the first quarter, with estimates pointing to increases of more than 50%.

“As customers anticipated further increases, actual contract prices came in higher, leading to the beat,” Kim Sunwoo, a senior analyst at Meritz Securities, said.

Samsung appears to be benefiting across most of its business. Analysts estimate its memory division generated about 54 trillion won in operating profit during the quarter.

Its mobile unit also held up, reporting around 4 trillion won in profit, though slightly lower than a year ago. However, its logic chip business is still under pressure and is expected to post a loss.

Currency movements have also helped. The South Korean won has fallen to a near 17-year low against the U.S. dollar, lifting the value of overseas earnings when converted back.

Even so, there are signs that the pace of growth may slow. The high cost of energy linked to the conflict in the Middle East has added pressure on production. At the same time, there are concerns that customers may begin to push back against high chip prices.

“There are growing concerns about a peak-out in memory price increases. It does appear that we are now past the initial upcycle phase and into a later stage,” said Ryu Young-ho, a senior analyst at NH Investment & Securities.

Recent data support that view. Spot prices for DRAM chips eased last week, noting that buyers are struggling to keep up with current price levels.

New technology could also affect demand. Google recently introduced a memory-saving system known as TurboQuant, which may reduce the amount of memory needed for AI workloads.

Samsung has also been working to strengthen its position in high-bandwidth memory chips, which are used in advanced AI processors.

The company began shipping its latest HBM4 chips to Nvidia in February, narrowing the gap with its main opponent, SK Hynix. Still, these advanced chips account for less than 10% of its DRAM revenue, meaning most of the profit is still coming from standard memory products.

In the market, Samsung’s shares rose 1.8% following the earnings outlook, outperforming the index. Shares in SK Hynix also increased.

Despite recent challenges, Samsung’s stock is still significantly higher this year, building on strong gains recorded in 2025.

The company is expected to release full details of its first-quarter results on April 30.

The company disclosed the raise alongside updates on its growth, revenue, user base and partnerships with global investors and technology firms.

OpenAI said the round drew support from strategic partners including Amazon, NVIDIA and SoftBank, with continued backing from Microsoft. Other participants included firms such as Andreessen Horowitz, D. E. Shaw Ventures, TPG and T. Rowe Price Associates.

Part of the capital also came through bank channels, with more than $3 billion raised from individual investors.

OpenAI added that its shares will be included in exchange-traded funds managed by ARK Invest, expanding access to the company’s equity ahead of a potential public listing.

The company also expanded its revolving credit facility to about $4.7 billion, supported by a syndicate of global banks. OpenAI said the facility is still undrawn.

In its statement, OpenAI pointed to rapid growth in revenue and usage. It reported monthly revenue of $2 billion and weekly active users exceeding 900 million across consumer AI products, alongside more than 50 million subscribers.

The company said, “We are now generating $2B in revenue per month. At this stage, we are growing revenue four times faster than the companies who defined the Internet and mobile eras, including Alphabet and Meta.”

OpenAI also noted adoption across both consumer and enterprise segments. It said enterprise customers now account for more than 40% of revenue, with expectations that this could match consumer revenue by the end of 2026.

Usage across its tools is increasing, with the company reporting commendable engagement in search, which it said has nearly tripled over the past year. It also stated that its advertising pilot has generated more than $100 million in annual recurring revenue within six weeks of launch.

On the developer side, OpenAI said its APIs process over 15 billion tokens per minute, while its coding tool Codex now serves over 2 million weekly users, with usage rising quickly in recent months.

The firm also referenced its latest model, GPT-5.4, which it said is driving higher engagement across agent-based workflows and enterprise applications.

Compute capacity is foremost to its expansion. OpenAI said its infrastructure strategy now spans multiple cloud providers and chip partners, including Oracle, CoreWeave and Google Cloud on the cloud side, and additional hardware collaborations beyond its long-standing reliance on NVIDIA systems.

The company described its approach as building a wider infrastructure base to support demand, rather than depending on a single provider.

OpenAI also outlined its comprehensive product direction, referring to plans to integrate its tools into a unified AI system. It said it is working towards a combined platform that brings together ChatGPT, Codex, browsing and agent features into a single interface.

The company stated that its long-term aim is to make AI tools easier to use across both personal and workplace settings, while allowing developers and businesses to build on top of its systems.

OpenAI added that its growth shows a mix of consumer adoption, enterprise deployment, developer usage and compute capacity, which together support continued expansion of its products and services.

The $122 billion funding round is one of the largest private capital raises in the technology sector and enables OpenAI to scale further as it prepares for greater market developments.

That is a 36% year‑on‑year increase from 2025, making this one of the fastest capital expenditure (CapEx) booms in modern corporate history. 

So, let’s discuss. Is this exceptional AI CapEx surge cycle driving productivity in the economy, or are we inflating another technological asset bubble?

The AI CapEx Scale: What’s Happening Now

Across the largest tech firms, the hyperscalers and cloud giants, capital spending is now structural. Amazon, Google, Meta and Microsoft are expected to put hundreds of billions into new infrastructure in 2026, much of it dedicated to specialised computing clusters, advanced networking and data centre capacity. 

The focal point of this spending is not mere servers or office upgrades. It’s data centres built specifically for high‑power compute workloads, facilities optimised for parallel processing at scale. 

These require specialised hardware like GPUs and high‑bandwidth memory, and they draw massive amounts of energy. 

One recent example shows just how strategic these moves have become. Nebius Group signed a multi‑year deal with Meta Platforms worth up to $27 billion to supply dedicated AI computing capacity by 2027, a contract driven by extreme demand and limited supply for high‑performance computing systems. 

Productivity: What the Investment Could Bring

No doubt that enhanced computing capacity enables economic value. Faster processing, more reliable inference workloads, and greater cloud availability can drive:

• Higher labour productivity by automating routine tasks.
• Faster research and development cycles in sectors from healthcare to manufacturing.
• Lower costs for compute‑intensive services, once infrastructure matures and utilisation improves.

For context, the semiconductor industry, a cornerstone of this infrastructure build‑out, is forecast to approach nearly $1 trillion in sales in 2026, with AI‑specific chips maintaining strong annual growth. 

From a macro perspective, such CapEx adds directly to aggregate demand and GDP in the short term. Data centre construction, advanced chip manufacturing, and supporting supply chains all contribute to economic activity that wouldn’t exist without this cycle. 

Bubble Territory: Where the Risks Begin

But there are strong arguments that we are edging into asset inflation rather than productive investment.

First, the pace of spending vastly outstrips current revenue realisation in the economy. Many of these specialised facilities operate at negative operating margins early in their life, requiring ongoing funding before they generate sustainable returns.

Second, a lot of the valuations attached to tech infrastructure assets incorporate lofty future earnings expectations. If those earnings don’t materialise, because adoption slows or competition increases, we could face rapid repricing. 

We’ve already seen some tension in the market, with certain historic investment commitments being scaled back. 

Third, hyperscalers are relying more on external financing even as their own cash flows get tighter. That’s a classic hallmark of an investment boom that may not be fully backed by near‑term productive returns. 

Semiconductors and Data Centres: The New “Oil”?

The analogy of compute as “the new oil” captures two truths:

1. Dependency: Modern AI workloads require massive compute capacity, just as 20th‑century industry relied on petroleum.
2. Infrastructure bottlenecks: Scaling compute, even with unlimited capital, is limited by semiconductor supply, power delivery, and cooling technology.

Already, suppliers like TSMC have posted strong revenue outlooks, showing reliance on advanced chips across the industry.

In parallel, smaller specialist data centre operators, such as CoreWeave, have expanded at a rapid clip. CoreWeave now operates dozens of facilities globally and has become a major supplier for bespoke compute capacity. 

But then, this infrastructure is expensive and energy‑intensive. Many facilities find it hard to break even without long‑term contracts or guaranteed utilisation.

Investment Implications: Winners and Fragilities

From an investment standpoint, certain firms appear ready to benefit if demand holds:

• Nvidia is at the centre of the compute supply chain. Its recent San Jose GTC 2026 forecast shows at least $1 trillion in chip revenue by 2027, driven by demand for next‑generation chips at scale. 
• Other chip designers and foundries stand to gain from backlogged orders and long production lead‑times.
• Data centre REITs and infrastructure funds may see longer‑term cash returns as contracts mature.

On the risk side, overcapacity, falling prices for older hardware, and slower adoption outside of hyperscale use cases are still substantive challenges.

So, Growth Engine or Asset Bubble?

Standing here in March 2026, we see both sides.

On the productivity side, this spending wave is building infrastructure that will underpin major advances in how industries operate. It’s tangible investment in capacity, not just speculation in intangible assets.

On the asset inflation side, the pace and scale of spending go beyond today’s revenue reality. Markets have priced future growth aggressively, which increases the risk of repricing if adoption deviates from expectations.

Now, are we financing a foundation for long‑term productivity, or are we inflating the price of future earnings prematurely?

The stock rose by more than 5% in early trading, and investors reacted quickly after Nvidia’s chief executive, Jensen Huang, spoke at the company’s GTC developer conference in California.

“I want to thank Samsung who manufactures the Groq LP30 chip for us and they’re cranking as hard as they can,” he said.

He added that the chips are already in production and will be shipped in the second half of the year.

That single update has changed the mindset around Samsung’s chipmaking business. For years, its foundry division has faced challenges, posting heavy losses as it tried to compete with bigger companies. Now, there are new signs of recovery.

At the same event, Samsung displayed the Nvidia chips built on its 4-nanometre process. It also introduced its latest high-bandwidth memory, aimed at handling growing demand from AI systems.

This points to a good working relationship between the two companies, especially as demand for AI chips continues to grow.

Experts say the deal could help Samsung’s foundry unit move closer to breaking even, though there are still challenges. Demand in the mobile market is still weak, and high memory prices could limit profits in the near term.

Even so, the market response shows good reports. Samsung shares were up about 4.3% at one point during the session, after earlier hitting stronger results. The index also moved higher, but not by as much.

There is more to watch this week. AMD chief executive Lisa Su is expected to visit South Korea, where she will meet Samsung chairman Jay Y. Lee.

Her visit is also expected to include a tour of Samsung’s chip plant in Pyeongtaek and it would be her first trip to the country since becoming CEO in 2014.