What's happening to blobs since EIP 4844? Ethereum and L2 analysis

About EIP-4844 and Blobs

Context for Ethereum and Layer 2s

The landscape of Ethereum Layer 2 blockchains has been completely transformed since the deployment of the Dencun update on March 13, 2024. This update introduced EIP-4844, a new mechanism designed to solve a critical issue for Ethereum Layer 2s: the cost of data availability on the main layer.

Before Dencun, Layer 2s were required to write their transaction data directly into Ethereum blocks due to the lack of alternatives. However, these blocks were not designed for this purpose, leading to congestion on the Ethereum blockchain. In other words, fees increased not only on Layer 2s but also on Layer 1, creating a lose-lose situation for all participants.

The implementation of Dencun addressed this issue by introducing a new fee market independent of Ethereum’s blocks, specifically tailored to the needs of Layer 2s, significantly reducing their fees.

How Blobs Work

To delve deeper, Dencun’s idea was to propose a system with independent fees and temporary data availability, rather than the permanent availability found in traditional blocks. In cases of congestion, this new market would not affect Layer 1 fees, and vice versa.

These new spaces in Ethereum blocks, called "blobs," are currently limited to six per block (with plans to scale up to 16 blobs per block). Upon Dencun’s implementation, a target of three blobs per block was set. This represents the threshold at which, if all are filled, fees start increasing significantly—by approximately 12% per block.

This mechanism incentivizes Layer 2s to use blobs while staying under the limit, encouraging data compression as the three-blob threshold approaches and blob space becomes more expensive.

Some key details:

• A blob consists of 4,096 field elements, each measuring 32 bytes.
• The storage capacity for blobs in a block is currently 0.75 MB.
• Blobs have a lifespan of two weeks and are deleted after this period to avoid burdening the main chain.
• The blob limit per block is six, with a target of three. This limit is expected to increase over time.

---

The Impacts of Dencun on Layer 2s

Fee Reduction

The Dencun update had an immediate impact on Layer 2 fees, which were reduced by a factor of 100 to 200. Fees dropped from an average of around $1 to less than $0.01 on some second-layer networks.

For Vitalik Buterin, Ethereum’s founder, this fee reduction was an essential step for the ecosystem’s adoption, enabling an affordable infrastructure for users while remaining highly profitable for Layer 2 sequencers.

In the short term, this drastic reduction in Layer 2 costs, coupled with narratives around airdrops and the rise of inscriptions (a trending phenomenon at the time), triggered a sudden spike in activity. The demand for the first three blobs surged before stabilizing.

However, one key observation stands out: the three-blob limit worked perfectly and was not exceeded, which was no coincidence.

The Blob Model Works

To understand this success, it’s important to remember that Layer 2s currently operate with centralized sequencers. In other words, they have the ability to manage certain actions themselves to optimize their expenses and revenues.

During the activity spike following the introduction of blobs, Layer 2s opted to delay “batching”—the process of grouping and submitting their transaction data to Ethereum’s main network. This batching process contains the entirety of a Layer 2’s activity over a given period, representing both data availability and security.

Generally, transactions are considered finalized and secure (recorded on Ethereum) when the sequencer completes batching. This step is therefore critical for users and decentralized applications.

Layer 2s Optimize the Blob System

A legitimate question arises: why would Layer 2s delay batching if it’s so important? The answer is simple: to maintain low costs by staying under the three-blob-per-block limit.

One of the main motivations is to reduce sequencer expenses, but it’s not the only reason. In most cases, revenues are directed to foundations (managed by DAOs), and Layer 2s also aim to remain competitive by offering the lowest fees to users.

Indeed, the cost of batching data on blobs is directly passed on to users. If the sequencer fails to optimize this process, Layer 2 users end up paying more, which could harm the network’s attractiveness.

This mechanism has its critics, with some questioning whether Layer 2s are fully aligned with Ethereum’s goals. Since Dencun’s implementation and the sequencers’ fee optimization, Ethereum’s revenue has dropped significantly. The idea that ETH could one day become deflationary again—through the burning of excess blob fees—has also been challenged.

---

Eight Months After Dencun: Key Takeaways

The Strategy Appears to Be Working

The Dencun update was part of a "loss leader" strategy aimed at offering blob space at a loss to compete with projects specializing in data availability, such as Celestia. The goal was to provide low-cost, highly secure data availability.

So far, Ethereum’s strategy appears to be working. At the time of Dencun’s launch, Ethereum had only around ten Layer 2s. Today, there are over fifty using blobs daily to store transaction data.

While not all of these Layer 2s have achieved the same success as Base or Arbitrum, and their per-block blob usage remains low, their sheer number creates sufficient scarcity.

Since late March 2024, blob demand has shown organic growth. This trend is further amplified by the creation of new Layer 2s, such as Unichain, Ink, and Worldchain.

Notably, networks like Taiko—a rollup inherently aligned with Ethereum—generate a significant demand for blob space to operate.

Will the Target Hold?

Recently, for the first time since Dencun’s introduction, Layer 2s have had to use more than three blobs per block. Unsurprisingly, this led to higher transaction costs. Why, then, did Layer 2s stop delaying their activities as they did in March 2024?

The answer is straightforward: Layer 2s cannot delay posting their transaction data to Ethereum indefinitely. As noted earlier, this would jeopardize user security. Moreover, since the data is public, users could notice the delay, leading to dissatisfaction.

This dynamic aligns with Dencun’s goal: when organic demand for blob space exists, exceeding the three-blob target becomes unavoidable. For instance, Optimism’s Superchain, which aggregates many Layer 2s, requires its Layer 2s to post to Layer 1 within 12 hours.

This recent breach of the three-blob target led to a noticeable increase in transaction fees, reaching levels not seen since Dencun’s implementation. This, in turn, reactivated ETH’s burn mechanism, albeit modestly.

---

What’s Next?

As mentioned, Ethereum aims to keep Layer 2 fees as low as possible. Therefore, the Ethereum community will likely vote to increase the target to four blobs, giving Layer 2s more breathing room.

Subsequently, as Layer 2s utilize this additional capacity, developers will introduce more blobs, increasing the limit to eight, then eventually reaching the maximum of 16 blobs.

Think of a port. If it has too few berths for a large number of ships, berth rental costs will skyrocket. Conversely, too many berths for too few ships devalues them, while still incurring maintenance costs. The goal is to balance supply and demand by gradually increasing capacity.

Ethereum will continue balancing profitability and non-profitability on blobs, gradually raising blob limits in response to demand until reaching the 128-blob cap outlined in the roadmap. At this stage, Ethereum is expected to become profitable solely through Layer 2 usage.

A simulator allows us to quickly test this scenario:

With 100 L2s achieving an average of 24 TPS and 64 blobs (i.e. half the roadmap), Ethereum would already become deflationary, without even including L1 activity, which remains stable today.

---

Conclusion

Despite recent doubts about Layer 2 activity and data availability competition, Ethereum’s blob system is showing strong promise. Incentives for creating Layer 2s using blobs appear robust enough to support the growth of an entire ecosystem around them.

With Dencun only eight months old, it’s too early to definitively judge its success. However, current trends suggest the roadmap is on track. Achieving its vision will require time, innovation, and additional blobs. Gradually, blob space could become a rare resource attracting significant demand.