The Dencun hard fork, implemented in March 2024, introduced 'blobs' for Layer 2 (L2) solutions, enabling them to post batched data to Layer 1 (L1) at significantly reduced costs.

As a result of this innovation, Base's operational expenses plummeted from $9.34 million in the first quarter of 2024 to just $42,000 by the third quarter.

Blob space is approximately ten times cheaper than L1 blockspace, making it a crucial component of Ethereum's rollup-centric roadmap.

This situation presents a complex challenge for Ethereum, necessitating careful consideration of future developments and their potential impacts.

In response to these challenges, Ethereum researchers are advocating for an increase in blob limits, proposing adjustments in the upcoming Pectra hard fork to prevent user exodus to cheaper chains.

Ansgar Dietrichs, an Ethereum researcher, acknowledged the misaligned incentives for L2s but emphasized the long-term importance of Ethereum's data availability.

Tim Robinson from Blueyard argued that while blobs currently generate minimal revenue, their design could lead to significant future benefits for Ethereum.

However, with the increase in on-chain activity during the current bull market, blobspace costs have begun to rise again.

Currently, each mainnet block can accommodate six blobs, with base fees kicking in when usage reaches three blobs and further increases at four.

Robinson's simulations suggest that an L1 processing 10,000 transactions per second with larger blob sizes could potentially burn 6.5% of ETH annually.

Despite the advantages of blobs, concerns persist among ETH holders regarding whether L2s will provide sufficient value to L1 over the long term.

Critics have raised concerns that L2s might opt for cheaper data availability solutions or delay posting data back to L1 if blob costs rise significantly.

Ultimately, the ongoing debate centers on whether to prioritize the average L2 user or the economic value of ETH, which could have significant implications for user retention and asset value.