Anthropic and AI: A Gigawatt Race to Dominate the Sector

A Record Fundraising Round for Anthropic

Dario Amodei, the CEO of Anthropic, recently announced a spectacular fundraising round of $65 billion, bringing the company's valuation to $965 billion. This operation confirms the market's appetite for leaders in artificial intelligence. However, beyond this impressive figure, a new economic reality is emerging for the AI sector: the growing importance of the physical resources needed to train and operate models.

For over a decade, technology has primarily focused on software. Today, artificial intelligence is reintroducing physical constraints, marking a shift towards an industry where hardware infrastructure is crucial.

Unprecedented Revenue and Valuation

With an annualized revenue now estimated at $47 billion, Anthropic positions itself in a category far beyond traditional startups. Few companies in the history of software have reached such a level of revenue so quickly. This performance partly explains the exceptional valuation of $965 billion granted by investors.

What distinguishes this fundraising round from previous ones is not only its amount but especially the intended use of the funds. The capital raised will not only be used to recruit researchers or develop new generations of models but also to secure the necessary infrastructure to meet a demand that is growing faster than available capacities.

Agreements for Ten Gigawatts of Capacity

Anthropic announced that it has entered into agreements representing up to five gigawatts of capacity with Amazon, along with an additional five gigawatts of next-generation TPU with Google and Broadcom. The company also benefits from access to the Colossus GPU infrastructure developed by SpaceX. The emergence of the gigawatt as a strategic unit likely indicates the ongoing transformation. Artificial intelligence is gradually ceasing to be primarily a software activity and is becoming a heavy industry. The situation resembles more the major phases of industrialization in telecommunications or cloud computing than the usual software cycles.

The Importance of Semiconductors

Another major takeaway from this operation is the presence of industrial partners like Micron, Samsung, and SK Hynix. These groups occupy a central position in the global supply chain of components used for artificial intelligence. Their participation illustrates a profound evolution in competition. For several years, the competitive advantage relied mainly on algorithms. Today, the availability of chips, HBM memory, and network infrastructure has become equally decisive.

The AI industry is gradually discovering the realities that industrial sectors have long faced: growth depends as much on the supply chain as on innovation itself. This evolution explains why memory manufacturers, component producers, and infrastructure operators are now playing an increasingly significant role in the AI lab ecosystem.

An Unprecedented Mix of Investors

The list of investors also illustrates this change. Alongside major venture capital names like Sequoia, Greenoaks, and General Catalyst, we now find Blackstone, Brookfield, Temasek, GIC, and Capital Group. This mix would have seemed improbable just a few years ago. The former traditionally fund technological innovation, while the latter typically invest in energy networks, transportation infrastructure, real estate assets, or essential equipment for the global economy. Their presence at the same table reflects how markets now perceive artificial intelligence.

A Global Trend

This dynamic is not limited to American giants. In Europe as well, AI players are gradually facing the same constraints. Mistral AI reminded yesterday of its intention to deploy 200 megawatts of capacity by the end of 2027, aiming for one gigawatt by 2030. Although the orders of magnitude remain distant from those mentioned by Anthropic, the strategic logic is identical.

The French company is no longer just looking to develop competitive models but also to secure the resources necessary for their operation and to build part of the European AI infrastructure. This evolution is particularly indicative of a change in era. Not long ago, the debate pitted American players with substantial resources against European players forced to focus on efficiency. Now, all are converging towards the same industrial reality.

A New Era for AI

Anthropic is not alone in this race. OpenAI is ramping up initiatives around Stargate to secure massive computing capacities. xAI continues to expand Colossus to support the development of Grok. Meta plans to invest several hundred billion dollars in its artificial intelligence infrastructure over the coming years.

The $65 billion fundraising by Anthropic thus marks an important milestone in the evolution of the sector. The first phase of AI was dominated by research, the second by the race for models, and the third is now focused on infrastructure. The most advanced labs are looking less and less like startups. They are securing energy capacities, negotiating long-term contracts with component manufacturers, reserving data centers, and building industrial alliances on a scale rarely seen in the software industry.