Zcash NU7 testnet cuts block time to 25 seconds before November decision

What happened

Zcash has switched on its next network upgrade on a public test network, weeks before a decision on whether the same rules should govern real funds. CoinDesk reported on October 5 that the upgrade, called NU7, activated at block 4,465,026 on the public testnet and cut the targeted wait between blocks from 75 seconds to 25 seconds. Software that follows Zcash and checks its transactions now lists the new rules as active on the public testnet, and an earlier, separate testing network has been shut down. The activation came ahead of the timing developers had previously estimated, which is why the test now has a longer public rehearsal than first planned.

A block records a batch of payments. Producing blocks more frequently can shorten the wait for a merchant to accept a payment or an exchange to credit a deposit. A service waiting for three confirmations would face a targeted wait of about 75 seconds instead of 225 seconds, if it kept its confirmation requirement unchanged. Actual block times vary around the target. The test network uses coins without monetary value, giving developers and infrastructure operators a place to check the changes before they affect real ZEC. Both major node programs now support the trial: one implementation was released October 1, and the Zcash Foundation published a test version of its node software the following day. Operators need compatible software to follow the upgraded test network. That software requirement is the immediate practical consequence of activation: a node that does not update cannot follow the new rules on the test chain.

The upgrade changes more than speed. NU7 also changes where transaction fees go. Miners receive 40 percent of fees, while the remaining 60 percent is removed from circulation and recorded in a reserve intended to fund future mining rewards. Those rewards pay miners to keep checking transactions as the amount of newly issued ZEC declines over time. Faster blocks do not triple the creation of new coins. The scheduled reward per block falls to one-third, preserving the overall issuance schedule. Block frequency, fee routing and per-block reward therefore move together as one design: more blocks, each with a smaller reward, and part of each fee set aside for future security spending.

The third change has a deadline attached for a specific group of holders. NU7 disables transactions used by Sprout, the oldest private-payment system in Zcash. Holders with funds in that pool would need to move them before mainnet activation to retain the ability to spend them through the existing system. No amount for funds still sitting in the legacy pool was reported in the coverage reviewed. The instruction for affected holders is still clear in the reporting: funds that are not moved before activation would lose their existing spending path. Testnet activation does not impose that deadline today. It starts the period in which exchanges, wallets and holders can prepare for it.

Developers plan to review the test results and decide on mainnet activation on October 20. November 5 remains the target for mainnet, with no mainnet activation block set yet. The sequence is deliberate: public test first, a dated review second, and only then a mainnet height if the review supports it.

Why it matters

The block interval is a user-facing property of a payment network. A target of 25 seconds instead of 75 changes how long services wait before treating a payment as settled enough to act on, assuming they keep the same number of confirmations. Exchanges crediting deposits, merchants accepting payment and bridge or custody systems watching for finality all calibrate to confirmation counts and block times together. A threefold faster target does not automatically mean every service will credit deposits three times faster, because a service can raise its confirmation requirement instead. The testnet period is when those operators decide which choice they will make, and their decisions, published in deposit and support notices, will show the real effect on user wait times.

The fee change matters for long-term security funding. Zcash, like Bitcoin, pays miners from a combination of newly issued coins and transaction fees. As issuance declines on its schedule, fees carry more of the security budget. Routing 60 percent of fees into a reserve for future mining rewards smooths that transition: fees paid during busy periods can support security during quieter ones. The reserve design also removes those coins from circulation until the reserve pays them out under the protocol rules. Readers should not read a price conclusion into that mechanism. The reporting establishes how fees are rerouted. It does not establish demand, fee volume in dollars, or what the reserve will hold by any future date. Those depend on network use that has not happened yet.

The Sprout change matters because sunsetting old systems is where upgrade risk becomes personal. Most holders use current wallets and current transaction formats and will notice only faster blocks. Holders with funds in the oldest privacy pool face a move-it-or-lose-the-path choice with a date attached, once the mainnet height is set. Networks that never retire old formats accumulate maintenance and security weight. Networks that retire them impose deadlines on the least active holders, who are often the least likely to hear about them. The quality of notice, wallet support and exchange communication between now and November 5 will decide how clean that retirement is. That communication record is a legitimate thing to watch, and it is visible in public support pages rather than in price charts.

There is also a governance point in the calendar. The mainnet decision is scheduled for October 20, after public test results can be reviewed, and the November 5 target has no activation block yet. That order, evidence first and height second, is the correct one for a change that alters block production, fee routing and transaction validity at once. A delay announced on October 20 with reasons would be a healthier outcome than an on-time activation with unresolved test failures. The testnet exists to make problems cheap. Its results should be allowed to move the date.

What to watch

The first marker is testnet behaviour under the new interval. Watch for published reports on block times in practice, orphan or reorganization rates, node synchronization problems and exchange or wallet test results. A target of 25 seconds is a setting. Sustained operation near that target, across many operators on compatible software, is the evidence. Problems on testnet are not a scandal. They are the purpose of the exercise, provided they are reported and fixed before real funds depend on the same code.

The second marker is the October 20 review. Developers are scheduled to decide on mainnet activation then. Watch whether the November 5 target survives, whether a mainnet activation block is set, and what conditions, if any, are attached. A decision that names remaining issues and a revised date would tell readers more than a simple confirmation. Either way, the mainnet date is not final until that review happens.

The third marker is Sprout migration support. Watch for wallet and exchange notices telling holders of the legacy pool how to move funds, deadlines stated in those notices, and any published figure for funds remaining in the pool as activation approaches. Clear, early notice with working tools is the standard to judge. Silence followed by a late rush would be the failure mode, and it would be visible in support channels before it shows anywhere else.

The last marker is fee and reserve reporting after any mainnet activation. Once NU7 rules govern real transactions, the reserve balance, fee share routed to miners and block interval in practice become measurable network statistics. Until then, the confirmed state is exact and limited: NU7 is live on public testnet at block 4,465,026, blocks target 25 seconds, 60 percent of fees are routed to a future-rewards reserve under the new rules, Sprout transactions are set to be disabled on mainnet, and a decision on the November 5 target is scheduled for October 20.