A landmark agreement between Nvidia and Amazon Web Services is offering one of the clearest signals yet of where the artificial intelligence economy is heading—and how the balance of power between chipmakers and cloud providers is evolving.

Nvidia will supply AWS with 1 million GPUs between now and 2027, according to the company’s vice president of hyperscale and high-performance computing, Ian Buck. The timeline aligns with chief executive Jensen Huang’s projection of a $1 trillion revenue opportunity tied to its next-generation Blackwell and Rubin chip architectures.

While the headline number is striking, the structure of the deal is more revealing. This is not a simple hardware purchase. It is a full-stack infrastructure partnership that spans compute, networking, and increasingly, inference—the stage of AI deployment where models generate responses and perform tasks in real time.

That distinction marks a turning point.

For much of the past two years, the AI boom has been defined by training—the process of building large language models using vast amounts of data and compute power. Nvidia’s dominance was built on supplying the GPUs required for that phase.

Now, the center of gravity is shifting toward inference. As AI systems move from development to widespread use, the demand profile changes. Instead of massive, one-off training runs, companies need sustained, efficient compute to serve millions—or billions—of user queries.

Buck captured the complexity of that shift bluntly: inference, he said, is “wickedly hard.”

To address it, AWS is not relying on a single class of chip. The deal includes a mix of Nvidia technologies—GPUs, Spectrum networking chips, and newer inference-focused processors such as Groq—alongside six additional Nvidia chip types. The goal is to optimize performance across different workloads, from large-scale model training to latency-sensitive applications like chatbots, recommendation engines, and autonomous systems.

This multi-chip approach reflects a broader industry reality. No single architecture can efficiently handle the full spectrum of AI tasks. Instead, hyperscalers are assembling heterogeneous compute stacks, combining different processors to balance cost, speed, and energy efficiency. That has implications for Nvidia’s long-term strategy. The company is no longer just a GPU vendor; it is positioning itself as a systems provider, integrating compute, networking, and software into a unified AI platform.

The inclusion of Nvidia’s ConnectX and Spectrum-X networking gear in AWS data centers is particularly significant. Traditionally, AWS has relied heavily on its own custom-built networking infrastructure, a core part of its competitive advantage. Opening that stack to Nvidia hardware suggests a deeper level of collaboration—and a recognition that AI workloads may require different architectural choices than traditional cloud computing.

It also signals a subtle shift in leverage. Hyperscalers like AWS have spent years developing in-house chips to reduce dependence on suppliers. Yet the scale and urgency of AI demand are forcing a more pragmatic approach: partnering with Nvidia even as they continue to build their own alternatives.

For AWS, the deal is about capacity and speed. Securing access to 1 million GPUs ensures it can meet surging customer demand for AI services, from startups building generative AI applications to enterprises embedding AI into core operations. But the agreement locks in long-term demand and reinforces Nvidia’s central role in the AI ecosystem. It also provides visibility into future revenue streams at a time when investors are closely watching whether the current AI spending boom can be sustained.

There is, however, a deeper competitive undercurrent.

The emphasis on inference chips—particularly newer offerings like Groq—suggests Nvidia is moving to defend its position against a growing field of specialized competitors. Startups and established players alike are targeting inference as a more cost-sensitive and potentially higher-volume segment than training.

If training established Nvidia’s dominance, inference will test its adaptability.

The economics are different. Training workloads are episodic and capital-intensive, favoring high-performance, high-margin chips. Inference workloads are continuous and cost-driven, requiring efficiency at scale. That shift could compress margins over time, even as total demand expands.

At the same time, the deal underscores the sheer scale of the AI buildout underway. A commitment of 1 million GPUs from a single cloud provider points to an infrastructure race that is still in its early stages. Data centers are being reconfigured, power consumption is rising, and supply chains are being stretched to meet demand.

This raises broader questions about sustainability—both in terms of energy usage and capital allocation. Hyperscalers are investing tens of billions of dollars in AI infrastructure, betting that demand will justify the outlay. Nvidia, in turn, is scaling production to meet that demand, tying its growth trajectory closely to the spending cycles of a handful of large customers.

The partnership with AWS illustrates how concentrated that ecosystem has become. A small number of companies—cloud providers, chipmakers, and large AI developers—are effectively shaping the architecture of the AI economy.

But AWS continues to develop its own chips, such as Trainium and Inferentia, aimed at reducing reliance on external suppliers. The coexistence of those efforts with large-scale Nvidia purchases reflects a dual strategy: build internally where possible, but buy externally where necessary to maintain competitiveness.

In that sense, the deal is both collaborative and competitive.

It locks Nvidia into the core of AWS’s AI infrastructure while reinforcing AWS’s role as a gatekeeper of AI services for enterprises. Each depends on the other, even as both seek to expand their own capabilities.

What emerges is a clearer picture of the next phase of the AI cycle. The initial scramble to build models is giving way to a longer, more complex process of deploying them at scale.

Goldman Sachs has warned that oil markets are entering a prolonged period of stress, with risks to prices tilted firmly to the upside as the U.S.-Israeli war with Iran disrupts supply flows and threatens to spill over into the global economy.

The latest escalation pushed Brent crude above $119 per barrel, underscoring the severity of the shock after Iranian strikes hit energy facilities across the Gulf. The attacks have forced production shut-ins and heightened uncertainty around the Strait of Hormuz, a critical artery that carries roughly one-fifth of global oil and gas supply.

Goldman’s analysis suggests the market may be underestimating how long the disruption could last. Drawing comparisons with past crises such as the 1973 oil embargo, the bank said supply shocks of this magnitude tend to persist, keeping prices elevated even after hostilities ease. While its base case assumes flows begin to recover from April and prices ease into the $70 range by late 2026, it warned that structural damage to infrastructure or prolonged geopolitical tension could keep oil above $100 per barrel into 2027.

That prospect is already reverberating across global financial markets, where the surge in oil prices is stoking fears of a broader economic downturn. Higher crude costs feed directly into fuel, transportation, and manufacturing expenses, raising input costs for businesses and eroding consumer purchasing power. For import-dependent economies, the shock is even more acute, often triggering currency pressures and widening trade deficits.

Economists say the current trajectory raises the risk of a stagflationary cycle—where growth slows while inflation accelerates. The longer oil remains elevated, the more likely it is to choke off demand, dampen industrial output, and weigh on global trade. Goldman noted that if disruptions persist, Brent could even surpass its 2008 peak, a scenario that would significantly tighten financial conditions worldwide.

Against this backdrop, central banks are being forced into a difficult position. The inflationary impulse from energy prices is complicating what had been a gradual shift toward monetary easing. Policymakers who were previously considering rate cuts are now reassessing their stance, with some weighing the need for tighter policy to prevent inflation from becoming entrenched.

The U.S. Federal Reserve, the European Central Bank, and other major monetary authorities are expected to hold a more hawkish tone in upcoming meetings, even as growth risks mount. Higher interest rates, while aimed at containing inflation, could further slow economic activity—deepening the risk of a synchronized global slowdown.

Goldman also highlighted the potential for additional market distortions. A widening spread between Brent and West Texas Intermediate could emerge if the United States considers export restrictions to shield domestic consumers, a move that would tighten global supply further. At the same time, while OPEC retains spare capacity, the bank cautioned that deploying it may not fully offset losses, particularly if infrastructure damage or security concerns limit production.

The crisis is exposing deeper structural vulnerabilities in the oil market. Years of underinvestment in upstream capacity, combined with rising geopolitical fragmentation, have reduced the system’s ability to absorb shocks. As a result, even partial disruptions are having outsized effects on prices and volatility.

Thus, the stakes are rising quickly for governments and policymakers. Elevated energy costs are not only a threat to growth but also a political risk, as households grapple with higher fuel and food prices. In emerging markets, where energy subsidies are often used to cushion consumers, fiscal pressures could intensify sharply.

In essence, the oil shock is no longer just a commodity story—it is becoming a macroeconomic one. With supply risks lingering, central banks on alert, and markets increasingly jittery, the trajectory of crude prices may now dictate the pace and stability of the global economic outlook in the months ahead.

Jeff Bezos is in early-stage talks to raise as much as $100 billion for an investment fund designed to buy manufacturing companies and overhaul them using artificial intelligence, according to the Wall Street Journal.

The proposed vehicle, described in investor materials as a “manufacturing transformation vehicle,” would focus on sectors such as semiconductors, defense, and aerospace. These are industries where production is complex, capital-intensive, and increasingly shaped by geopolitical priorities.

People familiar with the discussions told the Journal that Bezos has already engaged some of the world’s largest asset managers and held meetings with sovereign wealth funds in the Middle East in recent months. Those conversations highlight the scale of capital required and the type of long-term investors likely to back the effort.

By this move, Bezos is tagging along with a broader shift in the deployment of artificial intelligence. Much of the early investment cycle has focused on software and digital services. Bezos’ plan points to a second phase, where AI is applied to physical production systems.

The objective is not simply efficiency. It is to redesign manufacturing processes. AI has been largely touted to simulate production environments, optimize factory layouts, reduce defects, and shorten product development cycles. In sectors such as chipmaking and aerospace, even marginal improvements can translate into significant cost savings and strategic advantages.

The fund appears to be linked to a parallel initiative, Project Prometheus, which is focused on applying AI to engineering and manufacturing across industries, including automobiles and spacecraft. The startup is in discussions to raise up to $6 billion, according to the Journal, adding to the more than $6 billion it has already secured, as reported by the Financial Times.

Project Prometheus recently added David Limp to its board. Limp’s background in hardware and logistics suggests a focus on operational execution, not just software development. The company’s co-founders, Sherjil Ozair and William Guss, have not publicly commented on the latest fundraising efforts.

Together, the fund and the startup point to a vertically integrated approach. One entity would deploy capital to acquire industrial assets. The other would provide the AI systems to transform those assets. The model resembles earlier technology shifts where control of both infrastructure and software created durable advantages.

The timing aligns with a period of structural change in global manufacturing. Governments in the United States, Europe, and Asia are investing heavily in domestic production capacity, particularly in semiconductors and defense. Supply chain disruptions during the pandemic and rising geopolitical tensions have accelerated that trend.

Bezos’ outreach to Middle Eastern investors also rings a bell. Sovereign wealth funds in the region are seeking long-term investments tied to industrial diversification and technology. A fund of this scale offers exposure to both.

There are clear economic incentives behind the strategy. Manufacturing remains less digitized than other sectors. Productivity gains have been slower, and labor costs remain a major factor. AI offers the potential to improve throughput, reduce downtime, and enhance quality control.

However, analysts note the substantial risks involved. Unlike software, manufacturing transformation requires physical changes to factories, equipment, and supply chains. Returns are slower and depend on execution at scale. Integrating AI into production systems also raises workforce challenges, including retraining and potential job displacement.

There is also competition. Large industrial companies are investing in automation internally. Governments are attaching conditions to subsidies, particularly in strategic sectors like semiconductors. Private equity firms are increasingly targeting industrial technology, raising valuations for potential acquisition targets.

In addition, the scale of the proposed fund has been noted as another challenge. Deploying $100 billion effectively would require a steady pipeline of large transactions and the ability to integrate multiple businesses across regions and industries.

Bezos has pursued large, long-term bets before. At Amazon, he prioritized infrastructure investment well ahead of demand. At Blue Origin, he has taken a similarly patient approach to building space capabilities. The manufacturing fund appears to follow that pattern, focusing on long-duration returns rather than quick exits.

However, this time seems to be different. What distinguishes this effort is its scope. It is not limited to a single sector or technology. The move attempts to apply AI across the industrial economy, targeting the systems that produce goods rather than the platforms that distribute or sell them.

As AI-driven efficiency gains have been touted to lift margins and change competitive dynamics, particularly in industries where scale and precision are critical, Bezos’ move, if successful, is expected to reshape how manufacturing companies are valued and operated.

For now, the discussions remain preliminary. Key details such as fund structure, investor commitments, and acquisition strategy have yet to be finalized.

Bezos has not publicly commented on the plan.

However, this move denotes that the next phase of the AI economy is moving beyond code and into factories. And as he did in other sectors, Bezos is positioning himself to play a central role in that shift.