2 CRSi HVDC Server Deployments, $610 M US Operator Contract, 194 NVIDIA Blackwell Units, and 2 European Deals (2025 to 2026)

AI Power Infrastructure Risk, NVIDIA’s HVDC Shift and 2 CRSi’s $610 M Contract

The explosive growth in AI is creating an unavoidable data center power crisis, forcing the industry to adopt new architectures like 800 V High-Voltage Direct Current (HVDC). While large-scale server contracts validate this technological shift, they also introduce significant execution and counterparty risks, as seen with key enabler 2 CRSi.

The Unavoidable Power Bottleneck

The core constraint on AI expansion is no longer chip supply but access to power. This has forced a fundamental redesign of data center electrical systems to support next-generation hardware. The procurement of advanced servers is the leading indicator of this mandatory infrastructure upgrade cycle.

• The power density of modern GPU racks is approaching 250 k W, a dramatic increase from the 7 k W required by legacy CPU racks. This has rendered traditional low-voltage power distribution inefficient and impractical for AI factories.
• Data center electricity consumption is projected to consume up to 9% of total U.S. electricity by 2030. This has made energy efficiency a primary design criterion, driving the adoption of architectures like NVIDIA‘s 800 V HVDC, which promises up to a 5% end-to-end efficiency improvement.
• Before 2024, the industry focused on air cooling and standard AC power distribution. The post-2025 period is defined by a rapid transition to HVDC and advanced liquid cooling to manage the thermal and electrical load of AI accelerators.

2 CRSi’s Commercial Validation and Risk

As a specialized designer of high-density servers, 2 CRSi has emerged as a critical hardware enabler for the HVDC transition. Its recent contracts provide strong validation for its technology but are now overshadowed by questions regarding its financial reporting.

• 2 CRSi secured a landmark $610 million contract with a U.S. data center operator, including an initial order of $290 million, to supply high-density servers specifically designed for this new power paradigm. This represents a major commercial validation of the HVDC model.
• The company also announced a €110 million sale of its Godì Blackwell Ultra servers in Germany and a separate €140 million contract in February 2026, confirming strong demand in Europe for its AI-focused hardware.
• However, a June 2026 short-seller report from Grizzly Research accused 2 CRSi of fabricating revenues and misrepresenting partnerships. This has introduced substantial counterparty risk and threatens the company’s ability to finance and execute its large order backlog, potentially disrupting a key part of the HVDC supply chain.

$1.8 T Data Center Expansion, NVIDIA and 2 CRSi’s Infrastructure Investments

The capital expenditure required to support AI workloads extends far beyond GPUs, with a projected $1.8 trillion data center expansion by 2030 creating a massive, non-discretionary investment cycle in the underlying power and cooling infrastructure.

Data Center Cap Ex Projections

Industry-wide spending forecasts confirm that the largest investments are now flowing into the physical layer of AI factories. These figures highlight the scale of the market opportunity for suppliers of enabling power and cooling technologies.

• Total data center equipment and infrastructure spending reached $290 billion in 2024, a figure largely driven by the capital expenditures of hyperscale cloud providers who are the primary customers for advanced AI infrastructure.
• Projections show that operators are planning a $1.8 trillion expansion by 2030 to meet demand from AI and cloud computing, underscoring the long-term nature of this investment cycle.
• The cost of a single, state-of-the-art AI data center is now estimated to be at least €4–5 billion, reflecting the capital-intensive nature of building the required power and cooling infrastructure from the ground up.

2 CRSi’s Landmark Contracts

2 CRSi‘s recent large-scale procurement agreements serve as a concrete example of this investment trend, converting market projections into tangible revenue backlog. These deals are a direct result of the need to build out physical capacity for next-generation GPUs.

• In June 2026, 2 CRSi announced a €110 million (approx. $120 million) order for 194 of its Godì Blackwell Ultra servers, which are based on NVIDIA‘s latest architecture, for a customer in Germany.
• This followed a €140 million contract announced in February 2026 for servers dedicated to large-scale AI model training and inference.
• The most significant validation is the $610 million multi-year contract with a major U.S. cloud company, which establishes 2 CRSi as a key supplier in the world’s largest AI market.

Table: AI Data Center Infrastructure Investment

NVIDIA’s 3 Key HVDC Power Partners and 2 CRSi’s Supply Chain (2025 to 2026)

To de-risk its product roadmap, NVIDIA is orchestrating a full-stack ecosystem for its 800 V HVDC architecture, forging critical alliances with power infrastructure giants and specialized server designers to ensure its next-generation platforms can be deployed effectively.

NVIDIA’s Core Infrastructure Alliance

NVIDIA’s strategy extends beyond chip design to architecting the entire data center stack. The company has assembled a coalition of established industrial leaders to build out the power and cooling systems necessary to support its GPUs.

• NVIDIA has established formal partnerships with infrastructure giants ABB, Schneider Electric, and Delta Electronics to develop and deploy the components for its 800 V HVDC architecture.
• These partners are responsible for critical systems like switchgear, busways, and power distribution units (PDUs) that can handle the high-voltage, high-current demands of 1 MW AI racks.
• This ecosystem approach ensures that data center operators have a pre-validated, end-to-end solution, which accelerates deployment and reduces integration risk for the entire NVIDIA power strategy.

2 CRSi’s Role as Hardware Enabler

Within this ecosystem, specialized firms like 2 CRSi play a crucial role by providing the final hardware link in the chain: the servers themselves. These partnerships are essential for translating architectural concepts into commercially available products.

• 2 CRSi‘s $610 million contract with a major U.S. cloud operator makes it a key implementation partner for deploying NVIDIA‘s HVDC vision at scale.
• The company has also partnered with Chemours to qualify and deploy two-phase liquid cooling solutions. This technology is essential for managing the extreme thermal density of modern GPUs and is a core component of its server designs.
• These alliances demonstrate a shift from competing on individual product specifications to delivering integrated, full-stack solutions, which is now the key competitive differentiator in the AI infrastructure market.

Table: NVIDIA and 2 CRSi HVDC Ecosystem Partnerships

US vs. Europe, NVIDIA and 2 CRSi’s Data Center Power Deployments

Deployment of next-generation AI power infrastructure is unfolding differently across key markets, with the United States focused on hyperscale build-outs driven by tax incentives, while Europe prioritizes energy efficiency and sovereign compute capabilities.

The US Market: Scale and Incentives

The U.S. remains the primary market for large-scale AI infrastructure, driven by the concentration of hyperscale cloud providers and favorable state-level policies. Recent contract awards reflect this market’s appetite for cutting-edge hardware.

• The $610 million contract secured by 2 CRSi with a U.S. data center operator is a clear signal of the scale of deployments happening in North America. This region continues to lead in capital expenditure for AI.
• States like Ohio and Texas offer significant incentives, including comprehensive sales and property tax exemptions for data center equipment, power infrastructure, and cooling systems, attracting billions in investment.

The European Market: Regulation and Sovereignty

In Europe, the build-out is shaped by a dual focus on developing sovereign AI capabilities and adhering to strict energy efficiency regulations. This creates a strong market for highly efficient, high-density solutions.

• 2 CRSi, a French company, has secured major deals in Europe, including the €110 million contract for Blackwell-based servers in Germany, demonstrating the demand for high-performance computing within the EU.
• The European Commission is preparing a Data Centre Energy Efficiency Package that will include minimum performance standards. This regulatory pressure accelerates the adoption of efficient technologies like HVDC and liquid cooling, creating a favorable tailwind for suppliers.

Commercial Scale HVDC, NVIDIA’s 800 V Architecture and 2 CRSi’s Servers

Technologies once considered niche, including 800 V HVDC power distribution and two-phase liquid cooling, have rapidly matured to commercial scale, validated by their inclusion in multi-hundred-million-dollar server contracts for platforms like NVIDIA‘s Blackwell.

NVIDIA’s 800 V HVDC Architecture

NVIDIA‘s flagship power architecture has moved from a concept to a commercially backed standard. It represents a fundamental technological shift required to power the next generation of AI.

• The 800 V HVDC architecture is designed to power IT racks consuming 1 MW and beyond, with deployments slated to begin in 2027. It simplifies power distribution and reduces wasteful AC-DC conversions.
• Key benefits include up to a 5% improvement in end-to-end efficiency and a 30-40% reduction in copper infrastructure costs, directly addressing both operational and capital expenditure concerns.

2 CRSi’s Cooled High-Density Servers

The server technology required to house next-generation GPUs is now commercially available and being deployed at scale. This confirms the readiness of the hardware ecosystem to support the HVDC transition.

• 2 CRSi commercialized its Atlas 1.8 GG 2 PIC server in February 2026. This model leverages two-phase liquid cooling developed with Chemours, a technology essential for managing the thermal load of densely packed GPUs.
• The company’s servers are designed to be compatible with upcoming platforms like NVIDIA Vera Rubin, ensuring that the hardware is ready to support future generations of accelerators that will fully leverage the 800 V HVDC infrastructure.

SWOT Analysis: NVIDIA’s HVDC Ecosystem and 2 CRSi’s Critical Role

The strategic push for HVDC power architecture represents a significant opportunity to alleviate the AI energy bottleneck, but its success is contingent on managing supply chain constraints and the financial stability of pivotal partners.

Table: SWOT Analysis for the HVDC Power Transition

2 CRSi’s Credibility Test, NVIDIA’s HVDC Rollout, and What to Watch in 2026

The most critical near-term variable for the HVDC ecosystem is 2 CRSi‘s response to significant allegations of financial misrepresentation; the outcome will serve as a key signal for investor confidence and the supply chain’s ability to execute on its massive order backlog.

• If 2 CRSi fails to adequately address the short-seller allegations, watch for disruptions to its financing and ability to fulfill its $610 M U.S. contract. This could force NVIDIA and its customers to find alternative server suppliers, potentially delaying deployments. A key signal will be the company’s update on Project AETHER, expected around August 5, 2026.
• If the first large-scale 800 V HVDC deployments in 2027 demonstrate the promised efficiency and cost savings, watch for an acceleration of market-wide adoption. Real-world performance data will be the ultimate validation and will likely be a major focus at events like NVIDIA GTC 2026.
• If evolving EU regulations impose strict minimum performance standards for data centers, this could be happening: a bifurcation of the market where only the most efficient technologies like HVDC and liquid cooling are viable for new builds in Europe, strengthening the position of companies like 2 CRSi and NVIDIA‘s partners.

The questions your competitors are already asking

This report covers one angle of NVIDIA’s emerging power infrastructure ecosystem. The questions that matter most depend on your work.

• alternative suppliers for nvidia high density servers
• data center power equipment supply chain bottlenecks
• companies commercializing two-phase liquid cooling
• new European Union rules on data center power usage

This report does not answer these. Enki Brief Pro does.

Your question, your angle, your framework. SWOT, PESTL, scenario modelling. The same niche depth, built around the decision your work actually depends on.

Run your first brief in Enki Brief Pro