In late 2025, the Ethereum community quietly brought the Fusaka upgrade to a close.
Over the past year, talk of core protocol upgrades may have faded—but many on-chain users have felt one clear change: Ethereum L2 fees are getting cheaper.
Today, whether it’s a transfer or a complex DeFi action, gas often costs just a few cents—or less. While the Dencun upgrade and Blobs played a huge role, the activation of PeerDAS (Peer Data Availability Sampling) in Fusaka marks Ethereum's true departure from the era of "full download" verification.
In short, Ethereum’s long-term ability to support large-scale apps depends not only on Blobs, but also on the next step that PeerDAS enables.
1. What is PeerDAS?
To understand why PeerDAS matters, it helps to start with a key milestone in Ethereum scaling: Dencun (March 2024).
EIP-4844 introduced Blob-carrying transactions, allowing L2s to move data storage from expensive calldata to temporary blob storage.
This reduced rollup data costs dramatically, enabling cheaper and faster L2 transactions without undermining Ethereum’s security or decentralization—and it helped usher in a “low-gas” experience for users.
But there’s a hard cap on how many Blobs each mainnet (L1) block can carry—typically 3–6—because bandwidth and disk space are finite.
Under the traditional model, every validator—whether a data-center operator or a home node—must download and propagate the full Blob data to verify it.
That creates a dilemma:
- Increase Blob capacity (scale): data volume surges, home nodes get priced out, and the network trends toward centralization.
- Keep Blob capacity low (decentralize): L2 throughput stays capped and can’t meet future demand.
In short, Blobs solved “where to put data.” But as more rollups post data more frequently, bandwidth and storage pressure becomes a new decentralization risk.
If Ethereum keeps requiring full downloads, scaling will run into physical bandwidth limits. PeerDAS is designed to remove that bottleneck.
In one sentence: PeerDAS is a new verification approach that avoids full downloads, making it possible to scale from ~6 Blobs per block toward ~48 or more.
2. Blobs solve “where”; PeerDAS solves “how to store and verify at scale”
As mentioned above, Blobs solved “where data is stored”—moving it from costly calldata into temporary Blob space.
PeerDAS focuses on scaling storage and verification without overwhelming node bandwidth: not every node needs the full dataset for the network to be highly confident the data is available.
The idea is reflected in the name itself: Peer Data Availability Sampling.
Here’s a simple analogy: imagine a library receives a thousands-page encyclopedia (the Blob data). Under full verification, every librarian (node) would need a complete copy to keep it “safe.”
In practice, that would push out smaller operators over time—making the network less decentralized.
With PeerDAS and techniques like erasure coding, the “book” is encoded into many fragments. Each node stores only a random subset, instead of the entire dataset.
In theory, if the network can collectively provide around 50% of the pieces, the original data can be reconstructed with mathematical certainty.
This is the key shift: PeerDAS distributes the data-availability workload across many nodes, instead of requiring each node to carry the full load.
Source: @Maaztwts
Before Fusaka, Blob counts were effectively capped at 3–6 per block. With PeerDAS, that ceiling can be raised significantly—making targets like 48+ Blobs per block feasible.
When you transact on Arbitrum or Optimism, rollup data posted back to L1 no longer needs to be fully broadcast to every node—enabling scaling without node costs rising linearly with throughput.
Overall, Blobs + PeerDAS together form a more complete data availability (DA) approach—and a key bridge from proto-danksharding toward full Danksharding.
3. A new on-chain normal after Fusaka
In recent years, third-party modular DA layers (like Celestia) gained traction largely because Ethereum’s native DA was expensive. Their core premise was simple: Ethereum data availability costs too much.
With Blobs and now PeerDAS, Ethereum becomes both cheaper and highly secure: L2s can publish data to L1 at much lower cost, and Ethereum’s large validator set provides strong security guarantees.
For third-party DA solutions, this is a major challenge: Ethereum is reclaiming DA leadership and significantly reducing the space for alternatives to compete.
You might ask: This sounds very protocol-level—what does it have to do with my wallet, transfers, or DeFi?
The connection is direct. If PeerDAS rolls out smoothly, L2 data costs can stay low over the long term—so rollups are less likely to raise fees due to DA cost rebounds. Apps can design high-frequency interactions without wallets and DApps constantly trading off features vs. cost.
In other words: Blobs help make L2s affordable today, and PeerDAS helps keep them affordable tomorrow.
That’s why PeerDAS—though low-profile—remains a key milestone on Ethereum’s scaling roadmap. It’s the kind of infrastructure you benefit from every day, even if you rarely notice it.
PeerDAS shows how careful mathematical design (like sampling) can help blockchains handle Web2-scale data without unduly compromising decentralization.
By this point, Ethereum’s data highway has been fully paved. What cars run on it is now a question for the application layer to answer.
Let’s see what the next wave brings.