Mainnet szn never stops. Congrats to the @SuccinctLabs team! More proofs, less doubts!

☁️👉 ICYMI this week: @ethereum turned 10…and got a 100 MB/s DA layer. @eigen_da V2 is live on mainnet with THREE-digit throughput. 12.8× @Visa capacity. Built for verifiable apps, AI & the decentralized internet.

ZK fraud proofs meet cloud-scale DA. ⚡️ Congrats to the @conduitxyz, @boundless_xyz, and @risczero teams on OP Kailua going live. EigenDA support is on the way, bringing high-throughput, verifiable data to optimistic + ZK hybrids. Learn more ↓

Congrats to @eigen_da on this huge milestone! EigenDA is now the first DA solution to hit 100 MB/s throughput, with 6.7x more bandwidth and 60x lower latency. A major unlock for Mantle Network’s infrastructure as we scale towards becoming the Blockchain for Banking ⚡️

this momentum is unstoppable. last year, we fit all chains on EigenDA at just 15 MB/s. we're relentlessly pushing the boundaries of what's possible. once data is available, verification is effortless. that's the power we're unlocking. we're making it so damn available we can run NASDAQ and every major exchange. open finance for all

Updated my DA chart. These numbers are based on the data availability needs for MegaETH at ~20k TPS. We're leveraging the expanded capacity of EigenDA to a store additional data which will allow network nodes to sync with the block producer much, much quicker.

Congrats to @Eigen_Da on this big milestone. Supporting joint customers like @Mantle_Official and @Celo has been a privilege 🫡

As Ethereum turns 10, a decade of unstoppable infrastructure under its belt, it gets a 100 MB/s data layer. Built as an Ethereum-aligned L2, not a separate L1. @eigen_da V2 is now live at mainnet. Sustains 100MB/s, peaks at 124MB/s, with average 5s latency. This is how you build Distributed Durable Verifiable systems. Now developers can dream bigger and build better.

Having a lot of fun migrating MegaETH to EigenDA V2 and cannot be more excited. The new architecture is elegantly minimal and composable. The simplicity leaves ample room for further scalability and graceful upgrades. It's been an especially fun experience working with Bowen on the EigenDA side on all of our crazy benchmark experiments! @sreeramkannan and I had a few conversations on optimal DA architecture during the summer of 2022 after our DispersedLedger paper. We concluded that it's possible to build a hyper-scale DA fully decoupled from consensus. It's so exciting to see all the ideas being available in a production system!!!

As @eigen_da launches v2 with increase from 15MB/s to 100MB/s throughput, I want to dive deep into how EigenDA achieves this compared to v1. In v1 (original EigenDA): • When L2's sequencer sends a block to EigenDA, the disperser (within EigenDA) erasure codes blob into chunks (smaller blob pieces). • Chunks are sent to different storage operators. • Each storage operator processes their own set of chunks (smaller blob pieces), so there are no two operators who process the same chunks. • This approach enables horizontal scaling: more operators → more chunks can be processed → higher throughput → higher TPS. In v2 (new EigenDA): • EigenDA separates metadata (blob headers) and data (encoded chunks) communication. The disperser sends only blob headers to DA nodes, which validate payment and rate limits before requesting data payloads. • Data encoding is streamlined by standardizing blobs to 8192 chunks, making encoding stateless and less complex. Data is encoded once for all quorums, reducing redundancy. • LittDB (specialized DB) was designed for fast, reliable key-value storage on standard hardware. It forgoes data changes and complex transactions, focusing on sequential data expiration to support scalability requirements. @fuel_network and @aevoxyz are already live, but as I wrote in my previous article, EigenDA is not limited to rollups and crypto at all, it is also infra that supports data throughput needed for any software in general.