What are Blockchain Networks? Ethereum, Solana, and Beyond

On September 7, 2021, El Salvador made Bitcoin legal tender. Critics scrambled for the off switch — surely a government, or a regulator, or a frustrated billionaire could just shut it down. They couldn't. There was no server to seize, no company to subpoena, no CEO to pressure. Bitcoin kept producing a block roughly every ten minutes, exactly as it had every day since 2009, indifferent to the headlines.

That permanence comes from a single architectural fact: a blockchain isn't hosted anywhere. It runs on a network — thousands of independent computers, scattered across the planet, each holding a full copy of the ledger and each enforcing the same rules. To stop the chain, you'd have to stop every one of them at once. Nobody can.

This lesson is about that network: the nodes that make it up, how they stay in agreement without a boss, and why one network (Ethereum) feels nothing like another (Solana) even though both are "blockchains." Understanding network structure is the difference between a trader who knows why a transaction costs $8 on one chain and $0.01 on another — and one who just guesses.

A node is any computer running the blockchain's software and connected to its peer-to-peer network. That's it. No special hardware required for most of them — a node can be a rack server in a data center, a laptop in a bedroom, or a $35 Raspberry Pi humming in a closet. What matters is what it does: store data, validate transactions against the rules, and relay information to other nodes.

Not all nodes do the same job. The network is a division of labor:

• Full nodes are the backbone. Each one downloads, stores, and independently verifies every block and every transaction against the consensus rules. They trust no one — they check everything themselves. Bitcoin has over 10,000 publicly reachable full nodes; a current Bitcoin full node needs roughly 600+ GB of disk space.
• Validator / mining nodes are the ones that create new blocks. On proof-of-work chains, miners burn electricity to win that right. On proof-of-stake chains like Ethereum, validators lock up capital instead — 32 ETH per validator — and are slashed if they cheat.
• Light nodes (SPV clients) download only block headers, not the full chain. They trade some independence for the ability to run on a phone. Most mobile wallets are light clients.
• Archival nodes keep the entire historical state — every balance at every block ever — which is what block explorers and analytics tools query.
• RPC nodes expose an endpoint that apps and wallets call to read data and broadcast transactions. When your wallet "talks to the chain," it's usually talking to an RPC node.

Here's the puzzle that kept distributed computing stuck for decades: how do thousands of strangers, with no leader and no central database, all agree on the same list of transactions in the same order — when some of them might be lying? Solving that puzzle is what a blockchain network actually does. The mechanism is called consensus.

The flow is the same on most chains. You broadcast a transaction. It lands in the mempool — a waiting room of pending transactions that propagates node-to-node across the network in a "gossip" pattern, each node telling its peers, who tell their peers. A block producer (miner or validator) scoops a batch of those transactions, packages them into a block, and broadcasts it back out. Every full node independently re-checks that block — are the signatures valid? Do the senders have the funds? Does it follow every rule? — and only then appends it to its own copy of the chain.

The two dominant ways to choose who gets to produce the next block:

• Proof of Work (PoW) — used by Bitcoin. Miners race to solve a brute-force math puzzle; the winner earns the right to add the block. Security comes from raw energy expenditure — to rewrite history you'd need to out-compute the entire planet's mining power. Slow and expensive by design.
• Proof of Stake (PoS) — used by Ethereum since the 2022 Merge, and most newer chains. Validators put capital at risk instead of electricity. Misbehave and the protocol slashes — confiscates — part of your stake. Far more energy-efficient, and faster to reach finality.

Occasionally two valid blocks appear at nearly the same instant and the network briefly forks into two competing histories. The rule resolves it cleanly: nodes follow the chain with the most accumulated work (PoW) or the most attested stake (PoS), and the losing block is orphaned. This is why a Bitcoin transaction isn't truly "final" until a few blocks have stacked on top of it — each confirmation makes a reversal exponentially harder.

Bitcoin remains the hardest blockchain to attack. Over 19,000 reachable nodes enforce its consensus rules, and mining difficulty sits above 600 EH/s — more computational power than many countries consume. Throughput: roughly 7 transactions per second. Finality: ~60 minutes for the standard six-confirmation threshold. Fees average $1–5 in calm conditions, though they spiked past $60 during the 2023 ordinals frenzy. Bitcoin does one thing — store and transfer value — and does it with unmatched security.

Ethereum hosts most of crypto's financial infrastructure. The September 15, 2022 Merge swapped energy-intensive proof-of-work for proof-of-stake, slashing the network's electricity consumption by 99.95% in a single upgrade. Today, 930,000+ validators stake a collective ~30 million ETH to secure the chain. L1 throughput sits around 15–30 TPS with 12-second blocks and ~13-minute probabilistic finality. Gas fees range from $0.50 on a quiet weekend to north of $50 when a viral mint sends everyone scrambling for block space. With over $50 billion in DeFi TVL, Ethereum is the gravitational center of on-chain finance.

Solana optimized for raw speed above nearly everything else. Its proof-of-history mechanism timestamps transactions before they enter consensus, enabling ~4,000 TPS in production. Slot times: 0.4 seconds. Fees: fractions of a cent. (The 65,000 TPS figure on Solana's marketing pages is a theoretical ceiling that no one has hit under real-world conditions.) The tradeoff is concrete: seven network outages in 2022 alone, and a validator set of ~1,900 nodes running on hardware that costs upwards of $5,000/year — keeping the network smaller and more centralized than Ethereum's. Despite that, Solana found clear product-market fit for high-frequency DeFi, consumer apps, and meme-coin trading where sub-cent fees genuinely matter.

Arbitrum launched its mainnet August 31, 2021, and became the largest Ethereum Layer 2 by TVL within months. By compressing transactions and posting them to Ethereum via optimistic rollups, it delivers sub-second confirmation and fees of $0.01–0.10 — backed by Ethereum's full security model for settlement. One catch: native withdrawals to L1 involve a 7-day dispute window, though third-party fast bridges can shorten that for a small fee.

BNB Chain, Polygon, and Avalanche round out the tier below. BNB Chain runs on just 21 validators — Binance controls most of them — keeping fees at $0.05–0.30 at the cost of meaningful decentralization. Polygon expanded from a single sidechain into a multi-product scaling ecosystem with major brand partnerships (Starbucks, Nike, Reddit) and fees under $0.01. Avalanche's subnet architecture lets anyone launch a custom blockchain tuned for a specific use case, with ~1-second finality and growing institutional interest.

Layer 1 is the base blockchain — the system that handles its own security, consensus, and data availability with no external dependency. When a transaction settles on Ethereum L1, it's validated by 930,000+ independently operated nodes distributed across every timezone. Bitcoin, Ethereum, Solana, and Avalanche are all Layer 1s. They stand on their own.

Layer 2 networks execute transactions off the base chain, then post compressed results back to the L1 for final verification. The L1 acts as an arbitration layer: if an L2 operator submits fraudulent data, the proof posted on L1 exposes the lie, and protocol-level penalties slash their stake. Users get speed and low fees from the L2, while security traces back to the L1's validator set.

Two rollup designs dominate right now. Optimistic rollups — used by Arbitrum, Optimism, and Base — assume transactions are valid by default and provide a 7-day challenge window for anyone to submit a fraud proof if something looks wrong. ZK rollups — used by zkSync Era, StarkNet, and Scroll — generate cryptographic validity proofs that mathematically verify correctness before posting to L1. No dispute window necessary, but proof generation is computationally intensive and still being optimized.

You're probably already using an L2 without thinking about it. Base, Coinbase's rollup that launched August 2023, regularly processes more daily transactions than Ethereum mainnet. Your USDC on Arbitrum is the same USDC as on Ethereum L1 — same contract standard, same redeemability — you just pay $0.03 per swap instead of $8. Same security source. Radically different cost.

Vitalik Buterin articulated the blockchain trilemma in a 2017 blog post, and years of engineering haven't disproven it. A single blockchain can meaningfully optimize for two of three properties — security (resistance to attacks), decentralization (number of independent operators), and scalability (throughput and cost). Maximizing all three simultaneously remains an unsolved problem at the protocol level.

Ethereum chose security and decentralization. 930,000 validators scattered across every timezone. Virtually uncensorable. The price: 15–30 TPS on L1, with fees that spike to $50 when demand surges.

Solana chose scalability and speed. Sub-second finality. 4,000 TPS. Penny fractions for fees. The cost: fewer validators running more expensive hardware, and a track record of outages — seven in 2022 — that would be unthinkable on Ethereum.

BNB Chain chose raw throughput and cheapness. 21 active validators, most influenced by Binance. Fees stay under $0.30. The security model asks users to trust a single corporation with effective control over consensus — a proposition that runs counter to blockchain's stated purpose.

None of these tradeoffs is inherently wrong. Different applications need different properties. Trading high-value assets worth six figures demands maximum security; Ethereum or an Ethereum L2 is the obvious choice. Minting 10,000 sub-dollar gaming NFTs in an afternoon requires speed and negligible fees; Solana handles that better than anything else. The trilemma doesn't dictate a winner. It dictates fit.

The industry's emerging answer is layered architecture: a maximally secure L1 handles settlement while purpose-built L2s handle throughput. Ethereum plus its rollup ecosystem already processes more combined transactions per second than Solana, without giving up a single validator. The trilemma doesn't disappear — it just gets managed across layers instead of crammed into one chain.

Moving assets between blockchains requires a bridge — a set of smart contracts that lock tokens on the source chain and mint equivalent representations on the destination. Conceptually straightforward. Historically catastrophic.

On March 23, 2022, North Korean state-sponsored hackers drained $624 million from the Ronin bridge by compromising 5 of its 9 validator keys. Eleven days earlier, on February 2, the Wormhole bridge lost $320 million to an exploit in its Solana-side signature verification. Neither was a freak accident. Bridge contracts hold enormous pools of locked assets and present massive attack surfaces — a single vulnerability in input validation or key management can drain hundreds of millions in minutes.

Bridge security has improved since those watershed exploits. Multi-party computation (MPC) validation distributes trust across independent signers so that no single compromised key is fatal. Decentralized relayer networks eliminate single points of failure. Formal verification and longer challenge periods have closed the most glaring attack vectors. But bridges remain the weakest structural link in the multi-chain stack. When you route $10,000 through one, you're trusting the contract code, the validator set, and every upstream dependency between your wallet and the destination chain.

Practically: use battle-tested bridges (the canonical Arbitrum bridge, Across Protocol, Stargate) and avoid newer protocols with unaudited contracts. Better yet, skip bridging when you can. Platforms like GaiaEx accept deposits from multiple networks directly — send assets from whichever chain they already sit on and let the platform handle routing internally.

Capital doesn't live on one chain anymore. ETH sits on Ethereum mainnet. USDC migrates to Arbitrum for cheaper DeFi. SOL stays on Solana. A trader's portfolio ends up fragmented across networks by default, and any serious platform needs to meet users wherever their assets happen to be.

GaiaEx handles this by accepting deposits from Ethereum, Arbitrum, Solana, and other supported networks into a unified trading balance. Pick whichever network has the lowest fee at the moment — no manual bridging, no third-party router. Once funds arrive, the source chain becomes irrelevant. You trade spot and perpetual markets from a single account. Your USDC carries identical purchasing power whether it arrived via an Ethereum L1 transfer or a Solana SPL transaction.

Withdrawals work the same way in reverse: choose the destination network based on where the funds need to go and how much you're willing to pay in gas. Need assets on Ethereum L1 for a specific DeFi protocol? Withdraw there. Want the cheapest exit? Pull to Arbitrum and pay fractions of a cent. Cross-chain volume has grown every quarter since 2021, and GaiaEx's infrastructure was built for exactly that trajectory — deep liquidity and professional-grade execution, agnostic to which chain your assets started on.