What is Solana? It is a general-purpose layer 1 proof-of-stake blockchain designed by Anatoly Yakovenko in 2018, with its mainnet genesis block generated on 16 March 2020. In 2026 the network runs on Alpenglow, a consensus overhaul that went live in epoch 1064 on 21 October 2026 at roughly 13:36 UTC and targets about 150ms transaction finality, down from roughly 12.8 seconds under the previous TowerBFT design.
Why It Matters in Current Market Cycles
Sentiment is defensive. The Crypto Fear & Greed Index read 10 out of 100 on 16 November 2026, deep in Extreme Fear territory, and that backdrop shapes how investors read infrastructure upgrades. Consensus changes do not create demand on their own, but they do change the cost and risk profile of using a chain.
Solana's pitch has always been throughput at low cost. The network typically processes around 3,000 to 4,000 transactions per second, with a theoretical peak of 65,000 to 100,000 TPS, according to the Solana Research Institute's financial institution guide. Visa, for reference, averages 1,700 TPS with peak capacity of 65,000 TPS. For builders weighing settlement venues, that gap is the core argument, and Alpenglow is the strongest evidence yet that the argument is being engineered rather than marketed. The network also retired its long-running beta label, a milestone covered in our report on Solana dropping beta after more than six years.
What Is Solana's Alpenglow Consensus and How Does It Work?
Alpenglow replaces the voting mechanics that governed Solana since launch. Its voting protocol, Votor, shipped in Agave 4.3 and removed vote transactions from the block stream, sending votes directly between validators and aggregating them into certificates. Rotor, the component intended to replace Turbine for block propagation, is planned for a later release and is not yet scheduled.
The headline number is finality. Solana's own upgrade documentation states that Alpenglow targets roughly 150ms finality, a 99% reduction from about 12.8 seconds, and tolerates 20% malicious stake combined with 20% offline stake. Block time was cut separately from 400ms to 200ms across four stages completed in October 2026. The 150ms figure describes finality, not block production frequency, a distinction that matters when comparing published benchmarks.
Timelines in public research do not fully agree. A Solana Research Institute guide published 30 April 2026 described Alpenglow as expected later that year and cited 400ms slots with epochs of 432,000 slots, roughly 48 hours, and leader turns of four slots at 1.6 seconds. The solana.com upgrade page, updated in October 2026, reports the protocol already live with 200ms slots. Readers should treat the older parameters as the pre-upgrade baseline.
Execution is unchanged. The SVM, transaction format, programs, and fee mechanics behave as before, and "confirmed" and "finalized" now describe the same state.
How Do Solana Fees and Transaction Sizes Work?
Every transaction pays a base fee of 5,000 lamports, or 0.000005 SOL, per signature, plus an optional priority fee priced against the compute units requested. A single transaction may request up to 1.4M compute units. Size is capped at 1,232 bytes, or 4,096 bytes in the v1 format, and each transaction carries signatures plus a message of account addresses, a recent blockhash, and compiled instructions, as documented in the Solana core concepts reference.
Supply follows a fixed genesis allocation plus inflation. 500 million SOL were minted at mainnet genesis, with additional coins issued at a defined rate to pay validator rewards, slightly offset by burning a portion of transaction fees.
Head-to-Head Benchmark Comparison
The table below separates protocol-level parameters from payment-network benchmarks, using only figures published in the cited sources.
Key Benchmark Comparison (Metric | Solana, TowerBFT era | Solana, Alpenglow live (2026) | Visa benchmark)
- Transaction finality: Solana, TowerBFT era: ~12.8 seconds | Solana, Alpenglow live (2026): ~150ms target | Visa benchmark: Not disclosed in cited source
- Block time: Solana, TowerBFT era: 400ms | Solana, Alpenglow live (2026): 200ms | Visa benchmark: Not disclosed in cited source
- Typical throughput: Solana, TowerBFT era: 3,000 to 4,000 TPS | Solana, Alpenglow live (2026): 3,000 to 4,000 TPS | Visa benchmark: 1,700 TPS average
- Theoretical peak: Solana, TowerBFT era: 65,000 to 100,000 TPS | Solana, Alpenglow live (2026): 65,000 to 100,000 TPS | Visa benchmark: 65,000 TPS
- Base fee: Solana, TowerBFT era: 5,000 lamports per signature | Solana, Alpenglow live (2026): 5,000 lamports per signature | Visa benchmark: Not disclosed in cited source
- Fault tolerance: Solana, TowerBFT era: Not stated in cited source | Solana, Alpenglow live (2026): 20% malicious plus 20% offline stake | Visa benchmark: Not applicable
Security, Custody & Regulatory Considerations
Alpenglow's fault tolerance is stated as 20% malicious plus 20% offline stake, a different security envelope from the assumptions used in classical BFT literature. Institutions should map that number against their own counterparty and slashing assumptions rather than assume equivalence.
Custody remains an operational question, not a protocol one. Validator economics depend on the inflation schedule and priority fee markets, both of which shift with congestion. Compliance teams also need to track how finality changes affect settlement confirmation policies: a 150ms finality target compresses the window in which a transaction can be reverted, but it does not remove the need for application-layer confirmations. Our breakdown of SOL transaction time improvements against Ethereum in 2026 covers the comparative side.
Practical Takeaways
- Finality is not block time: Alpenglow targets roughly 150ms finality while block production runs at 200ms.
- Fees are predictable: 5,000 lamports per signature as a base, with priority fees tied to compute units up to 1.4M.
- Supply is genesis allocation plus inflation: 500 million SOL minted at mainnet genesis, with validator rewards partly offset by fee burns.
- Execution is unchanged: the SVM, programs, and transaction format behave as before Alpenglow.
- Verify parameters before citing them: older research still lists 400ms slots and 432,000-slot epochs, which describe the pre-upgrade baseline.
For builders, the practical question is not whether Solana is fast, but whether its fee market and fault-tolerance assumptions fit a given application. For investors, the Alpenglow rollout is a measurable engineering milestone in a market where sentiment, at 10 out of 100 on the Fear & Greed Index, is doing most of the talking.
Frequently Asked Questions
What is Solana in simple terms?
Solana is a general-purpose layer 1 proof-of-stake blockchain designed by Anatoly Yakovenko in 2018, with mainnet genesis on 16 March 2020. It typically processes around 3,000 to 4,000 transactions per second, with a theoretical peak of 65,000 to 100,000 TPS.
How fast is Solana after the Alpenglow upgrade?
Alpenglow went live in epoch 1064 on 21 October 2026 at roughly 13:36 UTC and targets about 150ms finality, a 99% reduction from roughly 12.8 seconds under TowerBFT. Block time was separately cut from 400ms to 200ms in four stages completed in October 2026.
How much does a Solana transaction cost?
Every transaction pays a base fee of 5,000 lamports, or 0.000005 SOL, per signature, plus an optional priority fee based on compute units requested, capped at 1.4M compute units per transaction.
Related Research
Primary Sources
- Crypto Fear & Greed Index — Alternative.me (Alternative.me)
- A Financial Institution's Guide to Solana | Solana Research (Verified Web Source)
- Core Concepts | Solana Docs (Verified Web Source)
Disclosure: This market analysis is provided for informational and educational purposes only and does not constitute financial or investment advice.