Four independent layer-1 networks — Mina, Solana, MultiversX, and VeChain — executed or scheduled hard forks within a 14-day window between September 3 and September 16, 2026. Collectively, the upgrades target order-of-magnitude improvements in finality, block time, and developer capacity. The cl...
"Block proposers can submit transactions before executing them, allowing validators to vote immediately while execution occurs asynchronously." — MultiversX Supernova Technical Documentation
Four independent layer-1 networks — Mina, Solana, MultiversX, and VeChain — executed or scheduled hard forks within a 14-day window between September 3 and September 16, 2026. Collectively, the upgrades target order-of-magnitude improvements in finality, block time, and developer capacity. The cluster is not coordinated; each chain is responding to distinct competitive pressures ranging from fee revenue collapse to EVM compatibility gaps.
The upgrades span different architectural strategies. Mina doubled its zkApp state capacity. MultiversX decoupled consensus from execution to cut block times 10x. Solana expanded transaction size 3.3x ahead of a full consensus replacement in October. VeChain leapt three Ethereum EVM versions in a single fork. Each approach reflects a different bet on what matters most for adoption: zero-knowledge programmability, raw throughput, payload flexibility, or tooling compatibility.
What unites them is timing. Layer-1 protocols that saw activity and fee revenue decline through H1 2026 are now shipping infrastructure bets that were conceived 12–18 months ago. Whether these upgrades translate into renewed usage or remain technical achievements without economic traction is the central question.
Mina Protocol activated its Mesa hard fork on September 3, 2026. The network halted transaction processing for approximately eight hours — five hours of blocks without transactions followed by roughly three hours of no block production.
What changed:
| Metric | Pre-Mesa | Post-Mesa | |--------|----------|-----------| | Block time | 180 seconds | 90 seconds | | On-chain state fields | 8 | 32 | | Event/action limits | Lower | Expanded (MIP8) | | Account updates per zkApp tx | Fewer | Increased (MIP9) |
The upgrade bundled four Mina Improvement Proposals (MIP6 through MIP9). MIP6 halved block production time. MIP7 quadrupled on-chain state fields from 8 to 32, directly expanding what zkApps can store on-chain. MIP8 and MIP9 raised event, action, and account-update limits per transaction.
A notable side effect: existing zkApps were temporarily paused. Mesa changed protocol constants and circuit constraints, invalidating pre-upgrade verification keys. Developers must redeploy verification keys before proof-authorized activity can resume. According to CryptoSlate, this effectively caused zero-knowledge applications to "go dark" post-upgrade until migration completes.
Mesa also introduced an automated hard fork mechanism intended to simplify future upgrades — an acknowledgment that Mina's prior upgrade cadence was too slow for the pace of development on competing chains.
MultiversX deployed the Supernova hard fork on September 10, 2026, during epoch 2233 at approximately 18:06 UTC. The network paused new transactions for roughly 24 minutes during the switchover.
What changed:
| Metric | Pre-Supernova | Post-Supernova | |--------|---------------|----------------| | Block time | 6 seconds | 600 milliseconds | | Intra-shard finality | Several seconds | < 250 ms | | Cross-shard settlement | ~24 seconds | ~2.4 seconds |
The core architectural change is the decoupling of consensus from transaction execution. Block proposers submit transactions before executing them. Validators vote immediately on block ordering while execution runs asynchronously. The resulting execution state is referenced and notarized in the following block.
According to MultiversX, 84.2% of mainnet nodes upgraded within one day of the release of v2.0.6.0, the final Supernova mainnet client. The upgrade was staged across testnet, devnet, and mainnet through the first ten days of September.
The 10x block time reduction is the most aggressive single-fork improvement among the four chains examined. However, MultiversX's total value locked and daily active addresses remain substantially smaller than Solana's or Ethereum's, limiting the immediate economic impact.
Solana activated Transaction V1 on mainnet September 9, 2026, defined by SIMD-0296 and SIMD-0385. The format had been live on testnet since epoch 1025 on September 1. Anza confirmed the mainnet date on August 29.
What changed:
| Metric | Legacy/V0 | V1 | |--------|-----------|-----| | Max serialized tx size | 1,232 bytes | 4,096 bytes | | Account limit per tx | Variable | 64 | | Instruction limit per tx | Variable | 64 |
The 3.3x size increase unlocks workloads that previously could not fit in a single transaction: zero-knowledge proofs, BLS signature verification, large multisig operations, and cross-chain message payloads. Legacy and V0 formats remain supported; applications that do not need the larger size are unaffected.
Transaction V1 is a precursor to Alpenglow, Solana's full consensus replacement. Alpenglow replaces Proof of History and Tower BFT with Votor (a new voting protocol) and Rotor (block propagation). Validator governance approved it via SIMD-0326 in September 2025. The Agave v4.3 release schedule names September 28, 2026, as the start of feature activation on mainnet, with full rollout targeted for October.
Alpenglow's headline numbers: deterministic finality in 100–150 milliseconds versus ~12.8 seconds today. Critically, on-chain vote transactions — currently ~75% of all Solana transactions — are eliminated entirely. Validators will exchange votes as lightweight UDP messages and use BLS signature aggregation to compress thousands of votes into compact certificates. The Solana Foundation states Votor tolerates 20% adversarial stake alongside 20% offline stake.
If Alpenglow delivers on its specifications, Solana's usable block space effectively triples overnight simply by removing vote transaction overhead.
VeChain's Interstellar hard fork (VIP-255) is scheduled for block height 25,902,540, expected at approximately 11:15 UTC on September 16, 2026. The community governance vote ran August 10–17, 2026.
What changed:
VIP-255 bundles 11 Ethereum Improvement Proposals in a single fork, advancing VeChainThor's EVM across three Ethereum release cycles — Cancun, Prague, and Osaka — simultaneously.
Key EIPs included:
VET surged 77% on the month leading into the upgrade. VTHO, the gas token, rose 145% over the same period, according to NullTX. The rally coincided with a new institutional validator going live on the network and VeChain publicly benchmarking its transaction architecture against Ethereum.
The EVM alignment strategy is functionally the opposite of Solana's approach. Where Solana is building a proprietary consensus stack optimized for raw speed, VeChain is collapsing the gap with Ethereum's tooling ecosystem. The bet is that developer portability and smart contract compatibility matter more than sub-second finality for enterprise adoption.
| Chain | Fork | Date | Downtime | Block Time (Before → After) | Key Change | |-------|------|------|----------|---------------------------|------------| | Mina | Mesa | Sep 3 | ~8 hours | 180s → 90s | zkApp state 4x, halved block time | | Solana | Tx V1 | Sep 9 | None | Unchanged (400ms) | Tx size 1,232B → 4,096B | | MultiversX | Supernova | Sep 10 | ~24 min | 6s → 600ms | Consensus-execution decoupling | | VeChain | Interstellar | Sep 16 | TBD | Unchanged (10s) | 11 EIPs, 3 ETH releases in 1 fork | | Solana | Alpenglow | Sep 28+ | TBD | 400ms → 400ms (finality: 12.8s → 150ms) | Full consensus replacement |
These upgrades arrive against a backdrop of declining layer-1 economics. Solana's case is the most documented: gross revenue fell from $1.09 billion in H1 2025 to $141 million in H1 2026, an 87.1% year-over-year decline, according to CCN. DeFi TVL dropped 22% quarter-over-quarter to $6.16 billion in Q1 2026, per Messari. Memecoin trading — which generated 95% of Solana's H1 2025 revenue through priority fees and Jito tips — fell from 40% to 16% of spot volume.
Mina, MultiversX, and VeChain operate at smaller absolute scales but face the same structural challenge: infrastructure capacity is expanding while the fee-paying user base is not growing proportionally. The upgrades increase throughput and developer capability, but throughput was not the binding constraint. Solana already processes 2,000–4,000 TPS; MultiversX had spare capacity before Supernova.
The economic-value question is whether faster finality and larger transactions attract new categories of on-chain activity — institutional settlement, real-world asset tokenization, cross-chain operations — or simply lower the cost of existing activity without generating incremental revenue. VeChain's EVM compatibility play is the most direct attempt to import existing demand rather than create new supply.
Four layer-1 hard forks landed in a 14-day window (Sep 3–16), with a fifth (Alpenglow) beginning activation Sep 28. This is an unusually concentrated period of base-layer infrastructure change.
MultiversX achieved the largest single-fork performance gain: block times dropped 10x from 6 seconds to 600 milliseconds, with intra-shard finality below 250ms.
Solana's Transaction V1 is a tactical prerequisite, not a standalone upgrade. The 3.3x size increase prepares the network for Alpenglow, which targets 150ms finality and eliminates 75% of current transaction volume (on-chain votes).
VeChain took the compatibility route, absorbing three Ethereum release cycles in one fork. VET rose 77% and VTHO 145% in the month preceding activation.
Mina's Mesa upgrade broke existing zkApps, requiring developers to redeploy verification keys. The trade-off: 4x more on-chain state and halved block times, but a forced migration for the existing developer base.
Revenue decline is the shared backdrop. Solana's 87% fee revenue drop is the most visible, but all four chains are shipping capacity expansions into soft demand environments.
Downtime varied significantly: Mina required 8 hours, MultiversX 24 minutes, Solana zero (for Tx V1). The range reflects different upgrade philosophies and network architectures.
September 2026 is the most concentrated period of layer-1 hard fork activity since the proof-of-stake migration wave of 2022–2023. The four chains are pursuing distinct strategies — speed (MultiversX), payload capacity (Solana), zero-knowledge programmability (Mina), and EVM compatibility (VeChain) — but share a common condition: infrastructure supply is growing faster than fee-paying demand.
The upgrades are technically substantial. MultiversX's 10x block time reduction, Solana's upcoming elimination of 75% of on-chain transactions, and VeChain's three-release EVM leap each represent years of engineering. Whether they generate economic returns depends on factors outside the protocol layer: regulatory clarity, institutional adoption timelines, and whether faster settlement translates into new use cases rather than cheaper execution of existing ones.
The data will become visible in Q4 2026 fee revenue, TVL trends, and developer activity metrics. Until then, these are supply-side bets awaiting demand-side validation.