Ethereum’s next leap is not about whether cryptography can prove a computation was correct. It can. The harder question is what still has to hold the whole thing together once the proof is done.
- Proofs verify correctness; they do not solve everything.
- PeerDAS improves data availability, not execution proofing.
- Ethereum still needs data access, ordering, inclusion, and censorship resistance.
That is the core tension in Vitalik Buterin’s September 27, 2026 essay, The cryptographic world computer. The idea is simple enough to sound almost dull at first: let nodes verify compact cryptographic proofs instead of re-executing every bit of computation themselves. Less waste. Less bandwidth. Less hardware pressure. Fine. Useful, even.
But a proof checking out does not mean the network is suddenly done. It does not mean data is available, transactions are fairly included, or censorship has been banished to the shadow realm where scam tokens belong.
A validity proof can show that a computation followed the rules. It cannot, by itself, guarantee that users can access the data, submit transactions freely, or know which result becomes final. That is the part the hype machine keeps trying to flatten into one sloppy slogan.
PeerDAS is live, and it solves a different problem
Ethereum’s December 2025 Fusaka upgrade brought PeerDAS to mainnet. PeerDAS stands for peer-to-peer data availability sampling, and the idea is simple: nodes do not need to download everything to check whether blob data is available.
Instead, they sample pieces of the data. If enough samples are reachable, the network can be confident the full data is available for reconstruction when needed. That matters because Ethereum has leaned heavily on rollups, and rollups depend on the base layer for data availability.
The Ethereum Foundation said Fusaka enabled an eightfold increase in theoretical blob capacity. In plain English, PeerDAS helps Ethereum move more rollup data without forcing every node to carry the whole load. That is the kind of scaling tradeoff Ethereum wants: more throughput, less pain, and fewer reasons for a home staker to feel like they need a datacenter in the garage.
According to Ethereum’s description, PeerDAS splits extended blob data into 128 columns for distribution and sampling, and regular nodes subscribe to at least 8 column subnets. The source notes that this is roughly one-sixteenth of the data by column count, with coding redundancy making the effective burden closer to one-eighth of the original data volume.
The practical takeaway is cleaner than the math: nodes sample, they do not hoard. That lowers bandwidth and storage pressure while preserving enough redundancy for the system to stay useful.
What PeerDAS does not do
Here is where the terminology gets messy, because crypto loves using one word to mean three different things.
Data availability asks whether the data exists and can be accessed. Validity proofs ask whether a computation was performed correctly. Ordering and inclusion ask who gets into the block, in what order, and whether anyone can be blocked along the way.
PeerDAS is about the first of those. It does not prove every calculation. It does not replace the need for transaction ordering. It does not stop a sequencer from acting like a gatekeeper. It does not, by itself, guarantee censorship resistance.
That distinction matters because “proof” gets used like a magic word in crypto marketing. It is not magic. It is a tool. A very good tool, but still just a tool.
Buterin’s essay puts the point plainly: “A proof can establish that a calculation followed specified rules, but users still need data, a way to submit transactions and a protocol that decides whose result becomes final.”
That sentence does a lot of heavy lifting, and it should. Because it is easy to confuse a correct computation with a live, trustless public network. They are not the same thing.
Who makes the proof, and who checks it?
The headline question has two answers.
The prover does the heavy work and generates the proof. The verifier checks the proof against public inputs and the protocol rules.
That split is the whole appeal of proof-based systems. If verification is cheap enough, ordinary nodes can still participate. The network can avoid making every participant redo the same expensive computation. That is the dream, anyway.
The catch is obvious enough once you say it out loud: someone still has to do the hard part. If proving becomes too expensive, too slow, or too specialized, the system can quietly drift toward a new kind of centralization. The verifier may stay open. The prover layer may not.
That is one of the real questions hanging over Ethereum’s future proof roadmap. Will proving remain broadly accessible, or will it end up concentrated among a small number of specialized operators with the right hardware and the right software stack? Crypto has a habit of promising decentralization and then rediscovering industrial-scale bottlenecks. Shocking, really.
Why Ethereum wants more than faster verification
Buterin’s long-term direction points toward a base layer that relies more heavily on cryptographic proofs for execution. A future zkEVM, or zero-knowledge Ethereum Virtual Machine, would let the network verify that block execution was valid without every node re-executing everything from scratch.
That is elegant. It is also only one piece of the puzzle.
Ethereum still has to preserve the properties that make a public chain worth using in the first place:
Data availability: users and nodes need access to the data so they can reconstruct what happened.
Transaction inclusion: people need a credible path to get transactions into the system.
Ordering: someone decides which transactions go first, and that power can be abused.
Censorship resistance: the network must still work when certain actors try to block access.
State reconstruction: nodes need enough information to rebuild the result, not just trust a summary.
Proofs help with correctness. They do not make those other problems disappear.
How this fits Ethereum’s rollup strategy
Ethereum has already made rollups central to its scaling plan. Rollups execute transactions off-chain and settle the results on Ethereum. The base layer provides data availability and final settlement, which is why blob data matters so much.
A validity proof can show that a rollup’s off-chain execution was correct. A validium also uses validity proofs, but does not post all transaction data to Ethereum mainnet. That can make it cheaper, but the tradeoff is brutal and simple: if the data is not onchain, users are trusting someone else to keep it available.
That is the sort of detail marketing copy tends to gloss over. A system can be cheaper and still be weaker. It can be valid and still be more trusting. It can be fast and still be a worse deal for users who actually care about independence.
In other words: not all proof-based systems are equally trust-minimized. A proof is not a virtue badge. It is just a proof.
The part that deserves skepticism
Ethereum’s proof-heavy direction is ambitious, and there is real technical value in it. But it also creates a new risk that people should not pretend away.
Proof generation can be expensive. That usually means specialized hardware, specialized software, and a tendency toward consolidation. If only a few actors can generate proofs quickly and reliably, they gain leverage. The verification layer may remain open while the production layer becomes quietly concentrated.
There is also the operational question. What happens if proof generation stalls? What if the prover fails to deliver on time? What if the rules change and client teams disagree about implementation details? What if the math is sound but the pipeline is brittle?
Those are not theoretical nitpicks. They are the kind of failure modes that show up when elegant design meets production reality. Crypto likes to celebrate the theorem and ignore the plumbing. The plumbing is where systems either survive or get embarrassed.
Roadmap markers are not guarantees
The source material also points to future milestones, including a planned Hegota fork in 2027 and a separate 2029 quantum target reported by Crypto.news. Those should be treated as markers, not promises.
That matters because roadmaps in crypto have a funny habit of becoming confidence theater. A named milestone does not mean the feature is finished. It means someone thinks it is worth aiming at.
Buterin’s essay should be read the same way: as a personal technical vision, not a final upgrade specification ratified by all Ethereum client teams. Ethereum is a distributed coordination problem, not a product launch calendar.
The real picture is more grounded. Fusaka is already live for data sampling. Ethereum is pushing toward more proof-based verification. But the network still needs the rest of the machinery, availability, ordering, inclusion, reconstruction, and resilient clients, before anyone should talk as if cryptography has solved decentralization once and for all.
Key questions and takeaways
-
Does a cryptographic proof mean the whole network is trustless?
No. A proof can show that a computation was done correctly, but users still need data access, transaction inclusion, ordering, and censorship resistance. -
What does PeerDAS actually do?
It lets nodes sample blob data instead of downloading all of it, which reduces bandwidth and storage pressure while preserving data availability. -
Why is data availability separate from execution correctness?
Data availability asks whether the data can be accessed and reconstructed. Execution correctness asks whether the computation followed the rules. Those are related, but they are different problems. -
Could proof generation become centralized?
Yes. Verification can be cheap and open while proving becomes expensive and specialized enough to concentrate among a small number of operators. -
Is Ethereum’s base-layer execution proof vision already live?
No. PeerDAS is live for data availability sampling, but base-layer execution proofs remain future work and are separate from existing rollup proofs.
The useful takeaway is not that cryptography replaces the network. It does not. The useful takeaway is that Ethereum is trying to separate the hard problems instead of pretending one neat proof can erase them all. That is more honest, more difficult, and much closer to reality.
Further reading
A few useful links on Ethereum’s scaling path, the mechanics behind PeerDAS, and the ongoing debate over what “scaling” actually means.
- Vitalik Buterin profile on Firefly Social
- Understanding the impact of climate change on global systems
- Fusaka is live: Ethereum scaling upgrade explained
- Decoupling data availability and execution in Ethereum
- Ethereum’s 110% address surge post-Fusaka: scalability win or hype bubble
- Vitalik Buterin’s plan to boost Ethereum scalability with zkEVMs and DAS
