Blackstart Digital acquires, modernizes, and operates brownfield infrastructure into AI-grade datacenter capacity — at a fraction of the cost and timeline of new construction, on sites with the power and connectivity that define the asset class.
AI was trained in the cloud. It runs where demand is. Blackstart builds the physical layer that puts compute where the world actually uses it.
The constraint on AI infrastructure is no longer chips or capital. It is energized, interconnected, power-dense real estate — and it is scarce.
We take buildings that already hold the infrastructure that takes longest to build — substantial power, fiber connectivity, structural envelope, strategic location — and convert them into purpose-built AI datacenters.
The asset class is defined by power. Greenfield development is gated by multi-year interconnection queues, permitting, and construction timelines the market cannot wait out. Hyperscale and enterprise AI workloads are scaling faster than new capacity can be energized.
Blackstart is built to find, secure, and modernize exactly the buildings that solve this — former research and technology campuses, telecom switching centers, carrier hotels, manufacturing facilities — and bring them online in 12–18 months versus the 3–5+ years required to build new.
Across our Silicon Valley flagship and target portfolio nationally, we acquire at 30–50% below replacement cost, retaining structural cost advantages that new entrants cannot replicate.
Adaptive reuse is not a cost strategy. It is an infrastructure strategy. The assets that take longest to acquire — power, fiber, location, structural envelope — already exist in the right buildings. The discipline is finding them.
Sites are selected on the strength of existing utility interconnection, behind-the-meter generation capability, and structural cost advantages that compound over the asset's life.
Carrier-dense fiber paths and strategic adjacency to network meeting points. Connectivity is acquired, not built — the rights of way and dark fiber that take decades to assemble.
Conversion designs target AI-class rack densities, with cooling architectures matched to inference and training workloads at the scale modern accelerators require.
Physical security, access control, and compliance frameworks built to the standards of institutional and government tenants from day one of operation.
Operational discipline modeled on the institutional datacenter standard: uptime reporting, transparent tenant relationships, and the operating muscle of a team that has run infrastructure at scale.
Adaptive reuse of capable existing structures is the single largest sustainability lever in datacenter development — avoiding the embodied carbon of new construction without compromising performance.
Three conditions converge to make the next several years the defining period for adaptive-reuse AI infrastructure. None of them is temporary.
Hyperscale and enterprise AI workloads are scaling faster than new capacity can be energized. Operators need capacity now, not on a five-year greenfield horizon, and will contract for speed and certainty.
Interconnection queues, permitting cycles, and construction timelines have made energized, interconnected capacity the binding limit on AI buildout. Power-ready sites command a durable premium.
An identifiable cohort of capable industrial, research, and telecom assets sits underutilized in the markets that need capacity most. The window to acquire them at value closes once the market re-prices.
Adaptive reuse is fundamentally a community story. The buildings we acquire are part of the economic fabric of the cities they sit in — and our approach is to bring them back into productive use, not redevelop them out of recognition.
Each conversion creates construction, modernization, and long-term operations employment, in partnership with local utilities, municipal stakeholders, and the surrounding community. The legacy of the building stays. The capability moves forward.
We operate to standards of transparency and engagement appropriate to the scale and visibility of the infrastructure we build, and to the public interest in how digital capacity is sited and run.
The team behind Blackstart is not a group of first-time founders. Across data center infrastructure, energy development, institutional finance, and enterprise revenue, the partnership has built, operated, and exited some of the most consequential platforms in the asset class.
32 years of internet infrastructure experience. Founded and served as CEO of Evocative (300 MW, 21-data-center portfolio) and Element Critical (250 MW national wholesale platform). Built two national data center platforms from inception through institutional exit.
Former Management Consultant at PwC. Founded Electriq Power (NASDAQ: ELIQ). Extensive data center and renewable energy development experience. Managing Partner at GSF, a family office focused on sustainable infrastructure.
25 years of technology and communications experience. PE/VC/CVC executive. Founded Amber and Public Wireless. Managed the CVP Family Office with AUM exceeding $600M. Led spin-outs at GE Ventures and SKT Ventures.
Asset management, finance, and energy development executive. Held senior positions at Ørsted, Sungevity, E.ON, and Galp across asset management, business development, and international expansion.
Over two decades driving revenue and strategic growth in the data center industry, with leadership roles at Digital Realty and Infomart. Delivered $147M+ in total contract value at Digital Realty, consistently exceeding targets up to 300% of plan.