Amperesand Secures US$80M to Build Faster Power Systems for AI Data Centers

With growing pressure on power infrastructure, Amperesand’s MV SST platform aims to fix AI’s energy bottleneck—starting with hyperscalers.

Amperesand, a next-generation power infrastructure company focusing on AI data centers and critical power applications, has closed an oversubscribed US$80 million Series A round, co-led by Walden Catalyst Ventures and Temasek. The funding signals strong investor confidence that power delivery—not chips or software—may soon become AI’s biggest growth constraint.

The round saw continued backing from Foothill Ventures, Material Impact, TDK Ventures, and Xora Innovation, alongside new participants SG Growth Capital, Acclimate Ventures, and Industry Ventures. Amperesand said the fresh capital will be used to speed up the rollout of its next-generation power infrastructure to support the rising energy demands of AI data centers globally.

The Power Problem Behind AI’s Growth

The rise of AI has created an unprecedented electricity demand. Data centers running large-scale AI training and inference models consume massive amounts of energy, and existing infrastructure is struggling to keep up. Industry lead times for critical components—transformers, switchgear, UPS systems—have quadrupled since 2019, now averaging 12–24 months.

This delay has created a global bottleneck: AI infrastructure may scale faster than the grids that power it.

Investors see this as a critical gap—and an opportunity.
As Young Sohn, Founding Managing Partner at Walden Catalyst Ventures, noted:

“AI is redefining the limits of computing. But without a revolution in power infrastructure, this growth simply can’t scale. Amperesand is tackling one of the biggest bottlenecks in data center power delivery by leveraging new high-power, semiconductor-enabled technologies for faster, more efficient, and more sustainable access to energy.”

Amperesand’s Answer: MV Solid-State Transformer

At the center of Amperesand’s strategy is its Medium Voltage Solid-State Transformer (MV SST) platform. It is designed to replace outdated transformer systems and deliver power more efficiently to high-density computing systems.

Key advantages include:

• Up to 80% smaller electrical footprint
• 10× faster time to power
• 50% reduction in installation labor
• Integrated hardware + software + AI-driven control systems

The platform unifies power electronics, intelligent monitoring, and AI-native controls in a single compact system. It also integrates built-in cybersecurity and regionalized supply chains, making it suitable for both on-grid and off-grid deployment — an increasingly important requirement for national security applications.

Commercial Traction & Hyperscaler Demand

Amperesand plans to deploy 30 megawatts of commercial systems in 2026, with hyperscale AI customers as a priority. Its first commercial units are scheduled for delivery to the Port of Singapore in partnership with PSA International. Multiple pilots with hyperscale data center operators have also been secured for next year, focused on accelerating deployment of next-generation GPUs.

Singapore’s selection as an early deployment hub is strategic. AI expansion plus limited land availability have put new pressure on energy planning — making high-density, low-footprint infrastructure a competitive advantage.

Building on Eight Years of R&D

Although it formally spun out of Nanyang Technological University in 2023, Amperesand is backed by eight years of research conducted through Xora Innovation. In 2024, it completed full-power and bidirectional testing of its Generation 2 MV SST system, paving the way for commercialization.

This year, it expanded engineering and manufacturing hubs in San Francisco and Reno, and expects to cross 100 employees by end-2025 as it begins full-scale deployment.

Strategic Positioning: Technology + Sovereignty

CEO and Co-Founder Brian Dow highlighted another layer of strategy: AI data centers are no longer just business assets—they are national infrastructure. Amperesand aims to provide governments and operators with secure supply chains, local manufacturing, and in-country data protection, making its platform a fit for both economic and security priorities.

Xora Innovation believes the company is poised for category leadership. Its Managing Partner, Phil Inagaki, noted that Amperesand has outpaced expectations in both power density and cost reduction — two crucial metrics for hyperscalers planning long-term AI infrastructure.

Conclusion — More Than a Startup, A Test Case for the AI Energy Era

Amperesand is not just solving a technical problem — it may be rewriting how AI infrastructure gets built. The next challenge for the AI industry is not computation, but energy delivery at scale. With AI projected to consume up to 10% of global electricity by 2030, power could become the next frontier of innovation.

If Amperesand can execute at scale, it could become a defining player in the AI energy transition — one that reshapes how data centers are built, powered, and secured in Asia and beyond.

Quick Takeaways (Short)

• Amperesand raised US$80M Series A, co-led by Temasek & Walden Catalyst.
• Its MV SST platform solves power bottlenecks in AI data centers.
• First 30MW commercial deployment planned for 2026, starting in Singapore.
• Infrastructure delays and power shortages make this a critical market opportunity.
• Investors expect Amperesand to lead next-gen power systems for AI at scale.