2025年欧洲深度科技报告（英）

THE 2025EUROPEAN DEEP TECH REPORTMarch 2025Executive Summary2Deep Tech is often misunderstood. It focuses on applying scientific and engineering breakthroughs to products for the first time.It is the origin of venture capital and has historically shaped regional sovereignty by catalysing transformative technology trends. Although more capital is typically required, most of it is invested in building IP moats rather than in advertising. Longer development timelines delay initial revenue, but once the technology matures, revenue growth often accelerates. While companies like Northvolt have made headlines, Deep Tech companies fail at rates comparable to others, albeit with a distinct risk profile that investors must understand. The exit landscape remains too immature but overall returns suggest that Deep Tech portfolios outperform conventional tech.1. DEEP TECH (MIS)CONCEPTIONSEurope has the potential to be a global Deep Tech hub. It boasts top-tier research with 6 of the top 20 universities and 9 of the top 25 research institutes. While its research landscape spans across a breadth of technologies, Europe lacks the entrepreneurial, risk-taking culture. Europe should focus on select centres of excellence like Oxford, Cambridge, ETH, EPFL or TUM. Additionally, while the US benefits from a strong founder flywheel with companies like SpaceX, Palantir, and OpenAI, Europe'sfounder ecosystem remains largely limited to shallow tech.2. THE EUROPEAN DEEP TECH OPPORTUNITY– VC Markets: Deep Tech provides a hedge against momentum investing; in 2024, €15bn flowed into Deep Tech—down -28% compared to 2021 highs—while regular tech fell by -60% versus 2021.– Countries: The 3 most important countries are the UK ($4.2bn), France ($3.0bn), and Germany ($2.7bn); Deep Tech accounts for one-third of all VC money, with London, Paris, and Munich, serving as the main hubs.– Stages: Although ample early-stage funding is available, 50% of growth capital still comes from investors outside Europe.– Exit: M&A activity has increased to $12.2bn, though the two largest exits—Darktrace and Exscientia—were publicly listed firms acquired by US players.3. FUNDING LANDSCAPEExecutive Summary34. SEGMENT DEEP DIVESNOVEL AI$3.0bn (+113%)FUTURE OF COMPUTE$1.2bn (-4%)SPACE TECH$1.0bn (+20%)NOVEL ENERGY$1.1bn (+75%)RESILIENCE$653m (+74%)COMPBIO & CHEMISTRY$500m (+59%)ROBOTICS$700m (+14%)Autonomous driving Wave $1.1bn Foundational models Mistral $500m Series B, Poolside, and DeepL $300m; with LLMs making up majority of AI fundingGrowth areas: Enhancing AI’s understanding of the physical world, developing agents, and applying AI to specialized verticals such as biotech and materialsQuantumQuantinuum $300m, Riverlane $60m Series CAI Inference Axelera AI $68m Series BPhotonicsGrowth Areas: Advancing AI inference (including edge computing) and deploying AI accelerators and photonic solutions to improve data centre energy efficiency and bandwidthIn-space transportation Exploration Company $160m Ser