Solana Observatory for Network Decentralization Analysis

The Solana network,
mapped and verified.

Multi-source geo-verified analysis of validator distribution, infrastructure coverage, and network health.

Epoch 949 41.47% complete Snapshot: Tue, 31 Mar 2026 18:41:18 UTC

Network Health

Current state of the validator set and cluster performance.

Active Validators
783
of 796 total, 13 offline
Delinquent Stake ? Percentage of total stake held by validators that are behind on voting. Below 1% is healthy.
0.03%
of total active stake
Skip Rate ? Stake-weighted average percentage of assigned leader slots that were skipped. Lower is better. A healthy network stays below 1%.
0.07%
stake-weighted
RPC Nodes
182
public RPC endpoints
Top Client Versions
3.1.10 500 validators
3.1.11 149 validators
0.818.30111 18 validators

Decentralization

How distributed is stake across countries, networks, and individual validators?

Market Concentration (HHI) ? Herfindahl-Hirschman Index. Used in economics to measure market concentration. Below 1500 is competitive. Above 2500 means a few players dominate. We calculate this for countries and network providers.

Imagine all validator stake as a market. HHI tells you how monopolized it is. A score near zero means thousands of equal players. A score near 10,000 means one player controls everything.

1953 by country Moderately concentrated
1425 by network provider Competitive
Stake Inequality (Gini) ? Gini coefficient measures inequality. 0 means all validators have equal stake. 1 means one validator holds all stake. Used the same way economists measure wealth inequality in a country.

Think of it like income inequality, but for stake. A high Gini means a few validators hold most of the network's stake, while hundreds of smaller validators share the rest.

0.749 across all validators High inequality
Nakamoto Coefficient ? The minimum number of independent entities that together control 33%+ of stake. Named after Bitcoin's creator. Higher means more decentralized.
ASN ? One network provider (TERASWITCH) hosts enough validators to exceed 33% of stake alone. This is a known concentration risk in Solana infrastructure.
1
Critical
Country ? Two countries (US and Germany) together hold over 33% of stake. Geographic concentration creates regulatory and infrastructure risk.
2
Critical
City ? Two cities (Frankfurt and Amsterdam) together exceed 33% of stake. Most validators cluster in major European datacenter hubs.
2
Critical
Validator ? It takes 20 independent validators to reach 33% of stake. This is the strongest dimension - no single validator dominates.
20
Good

Infrastructure Coverage

Adoption of advanced networking infrastructure across the validator set.

DoubleZero
? A private fiber network that connects Solana validators, bypassing the public internet for faster and more reliable block propagation.
453
validators connected
56.9%
of all validators
47.2%
of total stake
95 devices 路 95 RPC nodes 路 159/164 links healthy
Jito BAM
? Block-level Attestation Mechanism by Jito. Coordinates validators for more efficient MEV handling and block building.
361
validators connected
28.4%
of total stake
13
BAM nodes
+ More sections in development: Validators Table, Per-validator History, Endpoint Monitor, Geo Discrepancy Analysis, Datacenter Pages, Historical Charts.

Geographic Distribution

Validator count by country. 15 countries represented.

馃嚛馃嚜 DE
185
32.6% stake
馃嚦馃嚤 NL
138
25.1% stake
馃嚭馃嚫 US
127
9.9% stake
馃嚡馃嚨 JP
68
6.5% stake
馃嚫馃嚞 SG
49
2.9% stake
馃嚞馃嚙 GB
41
9.2% stake
馃嚙馃嚪 BR
20
0.7% stake
馃嚤馃嚬 LT
16
4.5% stake
馃嚚馃嚘 CA
16
0.8% stake
馃嚟馃嚢 HK
15
0.5% stake

About SONDA

What is SONDA?

SONDA started as private scripts to cross-verify validator geolocation while choosing a datacenter for the Solya validator. Existing tools reported conflicting locations, so we built a multi-source verification pipeline instead.

The result is an open-source observatory tracking where validators actually run, how infrastructure coverage evolves over time, and what decentralization looks like in practice.

What makes it different?

  • 4-source geo cross-verification per IP (DB-IP, IPInfo, GeoJS, ip-api)
  • Discrepancy detection - flags validators where sources disagree
  • Full location history since 2021 across 785+ epochs
  • DoubleZero and Jito BAM integration with live data
  • Three clusters: mainnet, testnet, devnet
  • Open source - github.com/SolyaUk/sonda