Blockchains are not equally fast. Bitcoin processes roughly 7 transactions per second. Ethereum's base layer handles around 15 to 30. Solana claims throughput in the tens of thousands. The difference isn't marketing — it's architecture. Each speed decision comes with a direct tradeoff in decentralization or security. Understanding why requires looking at what actually limits throughput at the protocol level.
The Three Variables That Determine Speed
A blockchain's transaction throughput is constrained by three interdependent variables: block size, block time, and consensus mechanism.
Block size determines how much transaction data fits in a single block. Larger blocks hold more transactions but require more bandwidth and storage to propagate across the network. Nodes that can't keep up with propagation get left behind — which reduces the number of participants who can feasibly validate the chain.
Block time determines how frequently new blocks are produced. Faster block production increases throughput but raises the risk of orphaned blocks — blocks produced before the previous one has propagated to all nodes. Orphaned blocks waste computational work and can create brief forks that the network must resolve.
Consensus mechanism determines how nodes agree on the next valid block. More efficient consensus can reduce coordination overhead and allow faster finality — but often requires fewer or more trusted validators, which concentrates the network.
These three variables are linked. Optimizing any one of them typically forces a tradeoff in another. This is the practical version of the blockchain scalability trilemma.
How Different Chains Make the Tradeoff
Bitcoin runs roughly 7 transactions per second. It uses 10-minute block times and a 1MB block weight limit (extended to roughly 4MB equivalent under SegWit). Bitcoin's design explicitly prioritizes security and decentralization — the slow, predictable pace is intentional. Running a full Bitcoin node is accessible to ordinary consumer hardware.
Ethereum's base layer processes 15 to 30 transactions per second under normal conditions. It uses 12-second block times and Proof of Stake consensus. Ethereum has roughly 900,000 active validators, making it one of the most decentralized proof-of-stake chains by validator count. The tradeoff for that decentralization is constrained base-layer throughput.
Solana claims throughput in the range of 65,000 transactions per second under optimal conditions, though real-world figures are substantially lower. It achieves this partly through Proof of History — a cryptographic clock that timestamps transactions before they're included in a block, reducing the coordination overhead between validators. The cost: Solana's validator hardware requirements are significantly higher than Ethereum's. The network has roughly 1,900 active validators. That validator concentration isn't a flaw in the design — it's a consequence of it.
Where Layer 2 Fits
Ethereum's scaling strategy doesn't involve making the base layer faster. It involves moving execution off-chain. Layer 2 networks like Arbitrum, Optimism, and Base batch thousands of transactions off-chain, compress them, and submit proofs or compressed data back to Ethereum's base layer for settlement. Throughput in the hundreds to thousands of TPS is achievable this way.
In 2024, L2 transaction counts exceeded Ethereum base layer transaction volume — the scaling strategy is working in terms of usage. But the tradeoff is complexity: users must bridge assets, and security depends on each L2's proof mechanism and the delay windows built into fraud-proof systems.
What's Changing
Ethereum's EIP-4844 (proto-danksharding), activated in March 2024, introduced a new data type called blobs — temporary, cheaper data storage specifically designed for L2 calldata. This reduced L2 transaction costs dramatically by lowering the cost of settling compressed transaction batches on Ethereum. It doesn't change base-layer TPS but makes the L2 path cheaper and more efficient.
Full danksharding — the complete version — would expand blob capacity by roughly 64x and remains several years away on the roadmap. The direction is clear: Ethereum's base layer is intentionally slow and secure; throughput lives at Layer 2.
Solana continues to push validator hardware requirements upward to sustain throughput. Network outages in 2022 and 2023 — caused partly by consequences of that design — have reduced in frequency as the team stabilized consensus. Whether that stability holds at scale remains an open question.
Confirmation Signals
- L2 transaction volumes exceeding Ethereum base layer volume (confirmed in 2024)
- Solana validator count remaining low relative to Ethereum despite growing network activity
- EIP-4844 deployment reducing L2 fees by 90%+ in the weeks after activation
Invalidation Signals
- A consensus mechanism that achieves high throughput without compressing the validator set — not yet demonstrated at scale
- A sharded base-layer architecture that maintains decentralization while increasing per-shard throughput substantially
Timing Perspective
The throughput gap between chains reflects structural architectural choices, not temporary engineering debt. Ethereum's base layer TPS is unlikely to change dramatically — the roadmap is explicitly L2-centric. Bitcoin's is even more stable; block size debates have been politically contentious since 2015 and no credible near-term change is on the table.
Solana's throughput claims are real under favorable conditions. The question is whether the decentralization tradeoff is a problem or a feature — and that depends on what you believe a blockchain is ultimately for.
What This Post Doesn't Cover
This explanation focuses on protocol-level throughput mechanisms. It doesn't evaluate which architecture will prevail or make predictions about adoption or token prices. The tradeoffs are real, and the right choice depends on use-case requirements that vary considerably across applications.
