One Month After Dencun: The Impact Of EIP-4844 on L2s and Ethereum
With Dencun and blobs live since March 2024, we were curious about the impact the change has on the Ethereum ecosystem and especially on Layer 2s.
So we dug into the data and did some analysis to find out how blobs are being used. In this article, we want to share a few charts from the dashboard we created for this exploration.
Let’s go 👇
How Blobs Are Used
Within the first month, ~365,000 blobs were posted to mainnet, representing about $2.6MM in fees. That’s ~48GB of ephemeral data. Blobs are being actively used by L2s, and despite the pre-launch talk of greater block latency, it’s now clear that blobs are being consistently included.
For the first two weeks, the number of blobs per block was well below the target of 3 blobs. Around the 27th of March, the number went much higher for a few days due to blobscriptions, but then stabilized around 3 blobs again.
L2 demand for blobspace is below 3 blobs per block, and users are taking advantage of cheap on-chain data. We expect this to be temporary, as L2s' willingness to pay should far exceed blobscriptions once there is more demand.
This can also be seen in the blob base fee. When blobscriptions started to dominate the blobspace, the price quickly adjusted with the EIP-1559-style pricing, but the base price is still pretty low:
As envisioned, this leads to greatly reduced costs for L2s. @rob_0x estimates the savings to be up to $74MM for the first month alone.
Blob Posting and Layer 2s
Looking at the number of blobs per block, it’s also clear that demand is not meeting supply.
Interestingly, about 60% of blocks have no blobs, but 20% of blocks have 6 blobs. We haven’t figured out why yet, but it seems that users either don't post at all or batch-post 6 blobs at a time.
If you look at L2s, they have different posting strategies.
Arbitrum, for example, posts based on the number of transactions on their network (the lows are at night when there's less activity).
Meanwhile, Starknet posts in batches about every 8 hours (after an initial trial period).
Summary
In summary, we can conclude that in the first month:
- 🎉 Blobs are actively being used
- 📊 Demand is currently lower than supply
- 💸 Savings for users are significant and real
- ⚙️ Different rollups have different blob-posting strategies
Feel free to explore the data yourself at dune.com/ephema/eip-4844-the-impact-of-blobs and let us know if you’ve got any questions ✨