Mentioned in This Article
1. What is the Zcash NU7 upgrade?
Zcash is scheduled to activate its NU7 network upgrade overnight on November 5 to 6 (UTC), cutting the target time between blocks from 75 seconds to 25 seconds. Developers describe it as the first step in a longer plan to make private ZEC payments work at much larger volumes, with later phases covering wallet privacy, proof verification and node storage.
The roadmap was discussed on a Bankless podcast episode titled The Complete Zcash Roadmap, featuring Zcash cofounder Sean Bowe and protocol contributor Dev Ojha, hosted by David Hoffman. The conversation outlined five upgrades, ranging from changes that arrive this week to ideas that remain conceptual. Scale is the common thread, because privacy-focused blockchains face costs that transparent chains largely avoid.
Zcash launched as a fork of Bitcoin's codebase with an emphasis on optional privacy, using zero-knowledge proofs that let a transaction be validated without revealing its contents. Upgrades are rolled out in numbered steps, and each one requires node operators, miners, exchanges and wallet developers to update their software before the activation point so that the network stays in agreement about the rules.
2. Why does scale matter for a privacy chain?
A transparent blockchain exposes every balance and transfer, so wallets, explorers and servers can answer questions about an address instantly and nodes can discard old data they no longer need. Zcash's shielded transactions conceal senders, recipients and amounts, which is the whole point of the design, but that secrecy removes many of the shortcuts other networks rely on. Everything from checking a payment to storing the history of spent coins becomes harder.
Ojha's stated long-term ambition is a network capable of 50,000 transactions per second. Reaching it would require changes in several layers at once: how quickly blocks arrive, how wallets learn about incoming funds, how nodes verify proofs and how much state they must keep. The five upgrades discussed map onto those layers, and each carries a different degree of readiness.
3. How do faster blocks change the network?
The first upgrade is the one nearest to activation. NU7 is the ninth major network upgrade in Zcash's history, and its headline change reduces block spacing from 75 seconds to 25 seconds, tripling the frequency of blocks. Issuance per unit of time is intended to stay the same, so faster blocks do not mean more ZEC entering circulation each day. The activation is tied to a specific block height, and developers have not ruled out delays if unresolved issues appear.
Because the size of each block stays the same, the amount of data moving around the network over time grows roughly in proportion to how often blocks appear. Ojha said bandwidth demand would rise to about three times its former level. He pointed to techniques used on other chains, namely erasure coding and improved gossip between nodes, as ways to move data more efficiently as throughput grows.
Zcash is keeping proof of work, the mining system in which participants compete to find the next block. Ojha framed the project as an attempt to deliver high throughput on a proof-of-work chain, something he said has not been done before. Faster blocks go live with NU7 this week, while the harder throughput engineering is expected to take roughly the next two years.
Mining randomness is why the average figure hides a wide spread. Because blocks are found by chance, some arrive much sooner than the 25-second average and others take far longer, so the network has to remain stable across that variation. Shorter targets make the spread more noticeable to users, which is one reason the engineering is described as demanding rather than a simple parameter change.
4. How could wallets learn about payments privately?
The second upgrade addresses a question Ojha posed directly: how does a wallet find out that it has been paid? On a shielded chain, a wallet cannot simply ask a server for its balance, because the server is not supposed to know which funds belong to whom. Historically, wallets have handled this by downloading and scanning every transaction, an approach that grows impractical as transaction volume rises.
The proposed solution is private information retrieval, known as PIR. With PIR, a wallet sends an encrypted query and the server computes an answer without learning what was asked, using homomorphic encryption, a form of cryptography that allows calculations on data that stays encrypted. Ojha said the workload reduces to large matrix-vector multiplications, which graphics processors now perform cheaply thanks to the boom in artificial intelligence hardware.
Hoffman raised the concern that such a system might reintroduce a trusted middleman. Ojha responded that wallets could switch among many providers, similar to the way Ethereum users choose among RPC providers, so no single operator would hold special power. Bowe added that Zcash has already used PIR in a recent coinholder vote.
The reference to AI is notable because it links two trends. The same graphics hardware demand that has reshaped computing markets lowers the cost of the mathematics behind PIR, which turns a theoretical privacy technique into something that can be offered to ordinary wallet users.
5. What is Tachyon, and how does it speed up verification?
The third upgrade belongs to Tachyon, a cryptography overhaul led by Bowe, the cryptographer behind the Halo proof system and the Sapling upgrade. One bottleneck Tachyon targets is proof verification. Today, full nodes check the zero-knowledge proof attached to each shielded transaction individually, which consumes computing time that rises with every additional payment.
Tachyon plans to use recursive SNARKs, which are proofs that verify other proofs, to aggregate many into one. A node would then check a single rolled-up proof instead of a long list. Ojha credited Ethereum with pioneering recursive proofs in production, and Hoffman compared the concept to zero-knowledge rollups on that network. The aggregation work is still in development: code is being integrated into the cryptography stack for the Zakura full node, with no mainnet activation and no public launch date.
Zero-knowledge proofs are central to Zcash. They let a sender demonstrate that a transaction follows the rules, for example that the coins exist and are not being spent twice, without disclosing the amounts or the parties. The trade-off is that generating and checking these proofs takes computation, and aggregation is aimed at reducing the checking cost that every full node bears.
6. How could the chain stop growing without limit?
The fourth upgrade tackles state bloat, the second bottleneck Tachyon addresses. Each shielded spend records a nullifier, a unique marker that prevents the same coin from being spent twice. Privacy chains must keep this ever-growing list indefinitely, unlike transparent chains, which can prune spent outputs and shrink their working data.
Under the proposed design, Ojha said, full nodes would hold roughly the most recent month of live state, while older history would remain queryable in a trustless way. Bowe credited the underlying technique to a paper he co-wrote with the cryptographer Ian Miers. Ojha cautioned that a network at 50,000 transactions per second could not run on a Raspberry Pi, but he suggested that node operators might need a few hundred gigabytes rather than something approaching a petabyte. Like proof aggregation, this idea remains in development as part of Tachyon.
The reason nodes cannot simply forget old nullifiers is that doing so would let someone spend a coin twice without detection. Any design that limits what nodes keep must therefore preserve the ability to prove, when needed, that a given coin has or has not been spent. That requirement is what makes the trustless queryable history an important part of the proposal.
7. Who would pay for private data access?
The fifth upgrade is the least developed and concerns economics more than engineering. On networks such as Ethereum and Solana, wallet companies like MetaMask and Phantom often fund the RPC servers that deliver data to users, partly because they earn revenue from those users. Zcash does not track its users, so only a handful of providers currently serve data for free, which depends on goodwill rather than a sustainable business model.
Ojha proposed a private prepaid subscription as a remedy. A user might buy 1,000 requests for a month, and the wallet would convert that payment into one-time tokens. Each query would spend a token, using the same double-spend prevention that protects ZEC itself. Providers could verify that a subscription is valid without learning which subscriber is making a request. Ojha described the problem as solvable, but it remains at the concept stage.
The design goal is that payment for access should not leak identity. A conventional subscription ties usage to an account, which is exactly the linkage Zcash users want to avoid. By turning prepayment into single-use tokens, the scheme would let a provider be paid and still remain unable to build a profile of any customer's activity.
8. What comes after NU7?
The immediate milestone is the mainnet activation of NU7 around November 5 to 6. Developers have said readiness work was still being completed in the weeks before, so the exact timing depends on the block height being reached and on no last-minute problem surfacing. Once the upgrade is live, observers will be watching how the network handles blocks arriving three times as often.
Beyond that, the harder work remains. Proof aggregation and the stateless design belong to Tachyon, which has no announced activation date, and the private subscription idea still needs a full specification. Bowe argued during the discussion that artificial intelligence makes it easy to track people across transparent ledgers, which he sees as the motivation behind the effort. The full episode also covers post-quantum plans, formal verification and governance for listeners who want more detail.
Anyone following the project should treat the longer-term items as plans rather than commitments. Timelines such as two years for the throughput work are estimates offered in conversation, and cryptographic changes of this size normally pass through public specification, review and testing before reaching the main network.