The news
Skeleton Technologies, based in Tallinn, Estonia, became a unicorn during the week of October 12–20, 2026, via a nine-figure investment from Siemens that valued the company above €899 million (approximately $1 billion). The funding will mainly go toward US manufacturing and Silicon Valley R&D, aimed at meeting hyperscaler demand for AI data-center power systems. The company plans to invest more than €179 million (roughly $200 million) in the United States, including domestic cell manufacturing. Skeleton is preparing for a US initial public offering in 2027.
The company makes supercapacitors — energy-storage devices that absorb sudden power spikes from GPU clusters — and says its technology makes AI data centers up to 40% more energy efficient. CEO Taavi Madiberk has argued that the binding constraint on AI is no longer how many GPUs you can buy, but how much compute you can extract from each megawatt of power.
What Skeleton does
Skeleton Technologies manufactures supercapacitors: devices that store energy electrostatically and can charge and discharge extremely fast. Unlike batteries, which store large amounts of energy but release it slowly, supercapacitors excel at handling short, violent bursts of power demand.
That characteristic maps directly onto the AI data center problem. GPU clusters do not draw power smoothly — they create sudden spikes as thousands of chips ramp up simultaneously. Those spikes stress electrical infrastructure, waste energy, and force operators to over-provision power capacity. Skeleton's supercapacitors sit in the power chain absorbing those spikes, smoothing demand, and — the company claims — improving overall energy efficiency by up to 40%.
Background and context
The investment question for founders is where the scarcest resource sits. The industry spent 2023–2025 discovering that GPUs were the constraint; 2026 is the year it discovered that megawatts are the harder one. Land with power access now trades at a premium, and "power availability" has entered boardroom vocabulary.
Skeleton's raise sits inside this wave. The same week, Rabot Energy raised a $50 million Series B with a model for turning homes into virtual power plants — a different end of the same thesis, that the energy system needs new flexibility to serve AI-driven demand. Together the two rounds show the theme's breadth: from Estonian hardware in the data center to German software aggregating residential batteries, capital is flooding every layer of the power-for-AI stack.
Siemens as the investor is strategically legible. The German industrial giant sells the electrical infrastructure that data centers are built from, and a stake in Skeleton deepens its position in the AI power chain. Corporate strategic investments of this size also function as commercial validation: Siemens would not put nine figures into a supplier whose technology its hyperscaler customers did not want.
The planned 2027 US IPO adds another dimension. European deep-tech companies increasingly list in the US to access deeper capital pools, and a unicorn valuation anchored by Siemens gives Skeleton a credible path. The €179 million-plus US manufacturing commitment also aligns with American industrial policy incentives for domestic energy manufacturing.
Why it matters for founders
For founders, Skeleton's unicorn moment carries three messages:
First, the AI value chain's scarcest layers are physical. While thousands of startups compete in AI software, the binding constraints — power, cooling, grid interconnection — are hardware problems with long lead times and deep moats. Founders with engineering backgrounds should look at where the AI buildout is physically stuck.
Second, strategic corporate capital can be more valuable than pure venture money at the growth stage. Siemens brings customer relationships with every hyperscaler, manufacturing expertise, and credibility no financial investor can replicate. Founders in deep tech should cultivate strategic investors early, not as a last resort.
Third, the CEO's framing — "the limit is compute per megawatt, not GPUs" — is a masterclass in narrative. It reframes the company's product from a component (supercapacitors) into the answer to the industry's defining constraint. Founders should practice describing their company as the solution to the customer's scarcest resource.
Angeled.com connects founders with investors.
Why it matters for investors
Skeleton's round confirms power-for-AI as a durable, multi-year investment theme rather than a momentary trade. For investors, the implications are:
The theme has room at multiple layers. Skeleton (hardware components) and Rabot Energy (virtual power plants) raising in the same week shows the opportunity spans equipment, software, grid services, and financing. Investors should map the full stack and look for unfunded gaps.
Watch the IPO. A successful 2027 US listing would validate public-market appetite for power-for-AI hardware and likely pull forward a wave of growth-stage financings. A weak one would signal that public investors still discount hardware relative to AI software — useful information either way for timing private investments.
Read Next
General Translation Raises $27 Million Series A Led by a16z to Build Agent-First Localization Infrastructure
LangChain Raises $300 Million Led by CapitalG at a $3 Billion Valuation to Build the Agent Stack
Mount Thor Raises $45 Million to Run Cloud Fleets of Macs for AI Agents
OpenAI Reportedly Raises $15 Billion at a $1.25 Trillion Valuation
Q-CTRL Raises $290 Million Series C Led by Salesforce Ventures, Reaching Unicorn Status
SE3 Raises $30 Million Series A Led by Accel



