A study published by the Cambridge Centre for Alternative Finance on March 12, 2026, analyzing 11 years of peer-to-peer network data against 68 verified submarine cable fault events, found that 72% to 92% of inter-country submarine cables would need to fail simultaneously before more than 10% of Bitcoin network nodes disconnect.
However, researchers Wenbin Wu and Alexander Neumueller determined that targeted attacks on key cable chokepoints or the top five hosting providers could achieve similar disruption with only 5% to 20% of infrastructure compromised—an “order of magnitude more effective” threat vector.
The first longitudinal study of Bitcoin’s resilience to physical infrastructure disruption provides empirical benchmarks for network robustness, revealing both remarkable tolerance to random failures and concentrated vulnerability to deliberate attacks on critical routing nodes.
The researchers applied a country-level cascade model to P2P network data spanning 2014 to 2025, running 1,000 Monte Carlo simulations per scenario. The critical failure threshold for random cable removal stands at 0.72 to 0.92—meaning 72% to 92% of all inter-country submarine cables would need to fail before node disconnection exceeds 10% of the network.
By contrast, a targeted attack on cables with the highest “betweenness centrality”—those serving as intercontinental chokepoints—drops the critical threshold to 0.20. Targeting the top five hosting providers by node count (Hetzner, OVH, Comcast, Amazon Web Services, and Google Cloud) requires removing just 5% of routing capacity to achieve equivalent disruption.
This asymmetry defines two distinct threat models: random failures that Bitcoin easily survives, and coordinated attacks that remain credible risks.
Over 87% of the 68 verified cable fault events studied caused less than 5% node impact. The largest single event—seabed disturbances off Côte d’Ivoire damaging 7-8 cables simultaneously in March 2024—knocked out 43% of regional nodes but affected only 5-7 Bitcoin nodes globally, approximately 0.03% of the network.
Cable events showed essentially zero correlation with Bitcoin price, registering a statistically insignificant correlation coefficient of -0.02. Infrastructure disruptions remain invisible against daily price volatility.
The study tracks how resilience evolved across four distinct periods:
2014-2017: Highest resilience, with critical failure threshold at 0.90-0.92, reflecting early geographic diversity
2018-2021: Sharp decline to 0.72 during peak mining concentration in East Asia, the network’s lowest point
2022: Partial recovery to 0.88 following China’s mining ban and forced redistribution
2025: Current threshold settling at 0.78
Geographic diversification of Bitcoin mining “has not materially altered infrastructure resilience,” researchers noted, as physical cable topology rather than hashrate distribution determines network connectivity.
The study revealed that Tor (The Onion Router) adoption by Bitcoin nodes “creates a compound barrier to disruption,” challenging assumptions that obscured node locations might hide fragility.
As of 2025, 64% of Bitcoin nodes use Tor, rendering their physical locations unobservable to researchers. The paper found that Tor relay infrastructure is concentrated in Germany, France, and the Netherlands—countries with extensive and redundant submarine cable and land border connectivity.
The researchers built a four-layer model to test Tor’s impact, consistently showing higher resilience than clearnet-only baselines, with Tor adding 0.02 to 0.10 to the critical failure threshold. Tor adoption surged following censorship events including Iran’s 2019 internet shutdown, the 2021 Myanmar coup, and China’s mining ban—demonstrating “adaptive self-organization” where community shifts toward censorship-resistant infrastructure inadvertently strengthened physical network resilience.
The top five hosting providers by node count—Hetzner (Germany), OVH (France), Comcast (United States), Amazon Web Services (global), and Google Cloud (global)—represent a critical vulnerability. A coordinated shutdown of these providers’ Bitcoin-related services could achieve significant network disruption with minimal physical infrastructure damage.
Certain submarine cables serve as intercontinental chokepoints with high betweenness centrality. Targeted severing of these specific routes, rather than random cable damage, dramatically reduces the failure threshold required to impact Bitcoin nodes.
According to the Cambridge study, between 72% and 92% of inter-country submarine cables would need to fail simultaneously before Bitcoin experiences significant node disconnection. Analysis of 68 historical cable fault events showed over 87% caused less than 5% node impact, with the largest event affecting only 0.03% of the global network.
Tor adoption increases Bitcoin’s physical infrastructure resilience. Despite making node locations unobservable—which some assumed could hide geographic concentration risks—Tor relay infrastructure is concentrated in Germany, France, and the Netherlands, countries with extensive redundant connectivity that are difficult to isolate. The study found Tor added 0.02 to 0.10 to the critical failure threshold.
Targeted attacks on either key submarine cable chokepoints or the top five hosting providers represent the most effective disruption vectors. While random cable failures require 72-92% infrastructure damage, a coordinated attack on cables with high betweenness centrality achieves equivalent impact with only 20% damage, and targeting major hosting providers requires just 5% routing capacity removal.