Simplified RowDAS Proposal: Ethereum Researcher Csaba Kiraly Cuts L2 Compute Burden 18x – A Game-Changing Step or Just Another Incremental Patch?

Ethereum nodes are still gasping for breath. The Layer2 scaling train keeps accelerating, but the data availability layer is where the real bottlenecks hide — compute cycles, storage pressure, and that constant reminder that full decentralization remains just out of reach. Then today, a single researcher's draft proposal drops like a quiet thunderclap: simplified RowDAS. Not the full revolutionary RowDAS with its row channels, but a stripped-down evolution of PeerDAS that promises to slash blob recovery compute work by 11 to 18 times in simulation. Csaba Kiraly did the math on 1,000 nodes and suddenly the numbers look almost too clean. 48.6 CPU-seconds down to 2.75. 91 down to 6.6. An 11x to 14x reduction. The code didn’t throw away the entire architecture — it just reallocated specific blobs to high-stake nodes that already hold the majority of columns, letting the rest of the network offload recovery work instead of doing it themselves. We didn’t expect the proposal to look this surgical.