New SpaceX IPO filings suggest frontier AI firms are beginning to treat compute infrastructure as a standalone commercial business, with Elon Musk's xAI agreeing to provide large-scale AI capacity to competitor Anthropic.
The filing disclosed that Anthropic agreed to purchase compute services delivered through xAI's Colossus and Colossus II AI infrastructure clusters through May 2029 under an agreement valued at roughly $1.25 billion per month. This arrangement is particularly notable because Anthropic competes directly with xAI in the market for frontier AI models and enterprise AI services, indicating that at least some AI developers are increasingly willing to buy large-scale compute capacity from rival infrastructure operators rather than rely exclusively on internally owned GPU fleets or traditional hyperscaler cloud platforms.
SpaceX also said in the filing that it "may enter into additional compute capacity agreements with third parties in the future," suggesting the Anthropic deal may not remain an isolated arrangement. Analysts said the disclosures point to a broader structural shift underway in the AI industry, where excess compute infrastructure itself is emerging as a monetizable strategic asset independent of the AI models running on top of it.
Compute as a strategic asset class
"This is less about excess capacity and more about compute becoming its own strategic asset class," said Sameh Boujelbene, vice president at Dell'Oro Group. "Frontier AI companies are building at a scale where infrastructure can be used both internally and commercially." The agreement reflects a fundamental evolution in how the industry views its massive data center investments. For years, companies like OpenAI, Google DeepMind, and Anthropic treated compute as a cost center tied to internal research. Now, with xAI's willingness to sell capacity to a direct rival, the market is beginning to view high-performance GPU clusters as tradeable commodities.
The shift has parallels to the early days of cloud computing when companies like Amazon began selling excess server capacity as AWS. However, the scale here is far larger. The $1.25 billion monthly payment implies an annualized contract value of $15 billion—more than many entire cloud regions generate. This puts a concrete price tag on frontier-scale compute, a figure that has remained opaque despite the industry's aggressive expansion. Analysts believe this transparency will help enterprises benchmark their own AI infrastructure costs.
More compute options for CIOs
For CIOs and enterprise infrastructure leaders, the disclosures may signal that AI infrastructure sourcing is becoming strategically more complex as the market evolves beyond traditional hyperscaler cloud consumption models. Shay Boloor, chief market strategist at Futurum Group, said that enterprises may increasingly source AI infrastructure from a broader mix of providers, including hyperscalers, neocloud operators, specialized infrastructure vendors, and even frontier AI labs themselves.
"The old assumption was that enterprises would simply buy AI capacity from the major hyperscalers," Boloor said. "This filing suggests the market is moving toward a more complex supply chain where compute can come from hyperscalers, neoclouds, frontier labs, vertically integrated AI platforms and specialized infrastructure providers." Boujelbene said enterprises should increasingly think of GPU infrastructure as both a sourcing and utilization challenge rather than simply a cloud procurement decision. "The key questions are no longer only 'which model should we use?' but 'where should workloads run, at what cost, and with what level of utilization?'" Boujelbene said.
The real challenge in AI deployments has been about accessing GPUs and managing them at scale affordably, said Arnal Dayaratna, research VP for software development at IDC. "Putting public price tags on these arrangements gives enterprises a clearer signal of what frontier-scale infrastructure actually costs, which is essential context for building realistic AI ROI models and understanding why inference costs, usage limits, and API pricing look the way they do. For CIOs, it also clarifies that the economics of AI services are set upstream of the software layer, largely before a vendor ever writes a line of product code."
Resemblance to cloud economics
Until recently, frontier AI companies largely treated compute infrastructure as a tightly controlled internal capability closely tied to proprietary model development. The SpaceX filing, however, suggests the economics of AI infrastructure may be evolving toward something more closely resembling cloud infrastructure markets, where compute capacity itself becomes commercially tradable. Boujelbene said the arrangement points to "more fluid compute-sharing models" emerging across the industry as infrastructure spending continues accelerating and AI demand remains high.
The filing repeatedly emphasizes the scale of xAI's infrastructure ambitions, referencing continued investment in "AI infrastructure, compute capacity, and power systems" needed to support expanding training and inference workloads. It also provides one of the clearest public reference points yet for the economics underpinning frontier-scale AI compute infrastructure, an area where pricing, utilization rates, and long-term return models have largely remained opaque despite the industry's aggressive datacenter expansion. Boloor said the agreement effectively places one of the first meaningful public market values on frontier AI compute capacity.
"The $45B Anthropic/SpaceX agreement shows that scarce, high-quality AI compute has become valuable enough that one frontier AI company is willing to pay another infrastructure operator tens of billions of dollars to access it," Boloor said. The disclosures, he added, begin putting "a dollar value around frontier compute capacity" while offering insight into "the pricing power of scarce GPU clusters and ROI for companies building these systems."
Analysts reject simplistic 'oversupply' interpretation
The filing has also fueled debate over whether the AI industry's aggressive datacenter buildout could eventually outpace enterprise demand for frontier AI services. But analysts cautioned against interpreting the Anthropic arrangement as evidence that major AI companies are sitting on large amounts of idle infrastructure. "I wouldn't frame this as clear evidence that frontier AI firms are overbuilding GPU capacity," Boloor said. "This is more of the natural evolution of AI compute becoming its own monetizable infrastructure layer."
He said frontier AI companies are effectively forced to build infrastructure ahead of demand because "training runs, inference demand and agentic workloads don't scale in a perfectly smooth line," while procurement lead times for GPUs, networking systems, memory, and power infrastructure remain lengthy. Alvin Nguyen, senior analyst at Forrester, similarly said the arrangement is likely to reflect the evolving workload dynamics rather than simple excess capacity. "There is enough demand for AI overall that all AI infrastructure is finding use," Nguyen said, describing the arrangement as "the natural evolution toward compute sharing and infrastructure monetization."
Historically, the semiconductor industry experienced similar cycles of overinvestment and subsequent leasing. During the dot-com boom, companies built massive server farms that later became hosting and cloud services. Today's AI infrastructure appears to be following a parallel trajectory, but with even higher capital requirements. A single cluster of Nvidia H100 or B200 GPUs can cost billions of dollars, making it essential for operators to find revenue streams beyond their own model training. The xAI-Anthropic deal validates this approach, potentially encouraging other frontier labs to explore similar arrangements.
For CIOs, the emergence of compute-as-a-service from AI labs offers both opportunity and complexity. On one hand, it provides an alternative to the hyperscaler oligopoly, potentially driving down costs through competition. On the other hand, it introduces new risks around vendor lock-in, data sovereignty, and reliability. Enterprises that rely on a lab's compute capacity must consider what happens if that lab shifts focus or suffers a supply disruption. Nevertheless, the trend toward infrastructure commoditization appears irreversible, as AI workloads continue to grow and diversify across training, fine-tuning, and inference.
The broader implications for the technology industry are significant. If compute becomes a standalone business, it could reshape the balance of power among AI players. Companies with the largest GPU fleets—like xAI, Meta, and Microsoft—could become not just model providers but infrastructure merchants. This could accelerate the development of open standards for interconnecting compute resources, similar to how cloud APIs standardized access to virtual machines. It may also spur new financial instruments, such as compute futures or capacity swaps, allowing enterprises to hedge against price volatility in GPU access.
From a regulatory perspective, the concentration of compute ownership raises antitrust questions. The same regulators who scrutinize cloud market concentration may need to examine whether frontier labs are using their infrastructure dominance to gain unfair advantages in model markets. The xAI-Anthropic deal, however, shows that even fierce competitors can coexist as buyer and seller, which may alleviate some concerns. Ultimately, the filing offers a rare window into the internal economics of AI at a pivotal moment. As the industry matures, the ability to separate compute from models will likely become a defining characteristic of the next phase of AI development.
Source: Network World News