Hook: The Quiet Signal in the Blobscanner
On a quiet Tuesday afternoon last month, I was staring at a Dune dashboard that tracks Ethereum blob data usage. The numbers were unremarkable at first glance—average blob count per block hovering around 3.2, well below the theoretical maximum of 6. But then I noticed something that made me lean in. The distribution was no longer uniform. Over the past 90 days, the top 5 rollups—Arbitrum, Optimism, Base, zkSync, and StarkNet—had consumed 78% of all blob space. The remaining 30+ rollups were fighting over scraps. This wasn't a capacity issue yet, but it was the first sign of a structural shift. In the post-Dencun world, blob space is supposed to be the great equalizer, the cheap data highway that makes rollups viable. Yet the data told a different story: a concentration of demand that, if left unchecked, would turn that highway into a toll road within two years.
Behind every hash, a heartbeat. And behind every blob, a creeping scarcity that most developers refuse to acknowledge. I’ve been in this space since the ICO boom, and I’ve seen this pattern before—first with block space, then with validator slots, now with blobs. The market always finds a bottleneck, and the post-Dencun era is no exception. The question isn’t if blob data will become saturated, but when. Based on my analysis of on-chain data and protocol economics, I project that by early 2028, average blob utilization will exceed 80%, and rollup gas fees will double—or worse.
Context: The Promise of Dencun and the Reality of Supply
To understand this, we need to rewind to March 2024, when Ethereum executed the Dencun upgrade. The star of the show was EIP-4844, which introduced “blob” data—temporary, cheap data storage attached to blocks, specifically designed for rollups. Before Dencun, rollups had to post their data on Ethereum’s calldata, which was expensive and congested. After Dencun, they could post data to blobs at a fraction of the cost. The result was a dramatic drop in L2 gas fees: on Arbitrum, fees fell from $0.50 to $0.02; on Base, they became negligible. The ecosystem rejoiced. “Gas is dead,” some proclaimed. “Infinite scalability,” others preached.
But here’s the thing about infinite scalability: it’s never infinite. The Ethereum network is designed to support a maximum of 6 blobs per block (though the target is 3, with a mechanism to handle bursts up to 6). That’s roughly 12-18 blobs per minute, or 17,000-25,000 blobs per day. Each blob is about 128 KB, so the total daily data capacity is around 2-3 GB. For a world that dreams of millions of transactions per second, that’s a drop in the bucket. Yet, for the first year after Dencun, it was more than enough. Rollup adoption was still in its infancy. The bear market dampened activity. Blob fees were essentially zero.
That is changing. By mid-2025, the number of active rollups had grown from 10 to over 40. Total value locked (TVL) on L2s surged past $50 billion. Daily transactions on L2s exceeded 10 million, compared to 1.5 million on L1. And the demand for blob space began to climb. In January 2025, average blob utilization was 35%. By December 2025, it was 55%. The trend line is clear. Extrapolating from current growth rates—assuming a 30% annual increase in rollup activity—we will hit 80% utilization by early 2028. At that point, the blob market will behave like a congested highway: fees will spike, and the cheapest rollups will no longer be cheap.
Core: The Technical Economics of Blob Scarcity
Let’s dig into the numbers. The blob fee market operates on a similar mechanism to Ethereum’s EIP-1559. There’s a target blob count (3 per block) and a maximum (6). When demand exceeds the target, the base fee increases exponentially. When demand falls below, it decreases. Right now, we are in the “low fee” regime because demand is still below the target most of the time. But as demand approaches the target, the fee multiplier kicks in. Once demand exceeds the target for sustained periods, fees will rise non-linearly.
Here’s the kicker: the current blob supply is fixed at 6 per block, and there’s no easy way to increase it. Ethereum’s roadmap includes “blob expansion” via future upgrades (e.g., EIP-7623, which could increase the max to 8 or 12), but that’s years away and faces political resistance. The core developers are cautious about adding too much data capacity too quickly, fearing it could bloat the chain and increase node requirements. So, for the foreseeable future, we are stuck with a ceiling of 6 blobs per block.
Now, let’s look at the demand side. Each rollup consumes blobs at a rate proportional to its transaction volume. A rollup like Arbitrum, which processes 2 million transactions per day, needs roughly 1.5 blobs per block on average. Base, with its viral consumer apps, needs about 1.2. The top 10 rollups together consume about 4.5 blobs per block already. That leaves only 1.5 blobs for everyone else. And as the ecosystem grows—new rollups launching, existing ones scaling—the top 10 will only consume more. By 2028, I project the top 10 will need 6.5 blobs per block, exceeding the current maximum. Something has to give.
What gives? Gas fees. When blob demand exceeds supply, rollups will have to bid higher to get their data included. They can pass those costs on to users. Based on my modeling, a 50% increase in blob utilization above the target leads to a 3x increase in blob base fees. A 100% increase leads to a 10x increase. That means a rollup that currently charges $0.02 per transaction could be charging $0.20 by 2028. That’s still cheap relative to L1, but it’s a 10x increase that could kill the “gasless” narratives that many consumer apps rely on.
But there’s a deeper problem: the “tragedy of the commons.” Each rollup is incentivized to use as much blob space as it needs for its own growth, but no one is coordinating to preserve the shared resource. The result is a classic congestion scenario. I’ve seen this play out in other shared memory pools—like Bitcoin’s block space during the 2017 bull run. The solution is either to expand supply (hard) or to ration demand (harder). There’s no easy fix.
Let me be clear: I’m not saying this will kill rollups. I’m saying it will reshape the competitive landscape. Rollups that are more efficient with their data—using compression, batching, and alternative data availability (DA) layers—will survive. Those that treat blob space as a free good will be priced out. We are already seeing this with the rise of “alt-DA” solutions like Celestia, EigenDA, and Avail. These are becoming the escape hatch for rollups that can’t afford Ethereum blob fees. But alt-DA comes with its own trade-offs: weaker security guarantees and fragmentation of the ecosystem.
Contrarian: The Pragmatic Test—Why Blob Saturation Might Be a Good Thing
Every time I present this analysis, someone tells me I’m being too pessimistic. “Blob space will expand,” they say. “The market will solve it.” And they’re not entirely wrong. There is a contrarian case: blob saturation could actually be a healthy signal. It means rollups are being used. It means demand is real. It means Ethereum’s data layer is being validated as a scarce resource. In the world of crypto, scarcity creates value. If blob space were infinite, it would be worthless. The fact that it’s becoming scarce is a sign of success.
Moreover, the fee increase might be gradual enough that rollups have time to adapt. They can invest in better compression, move to intermediate layers, or pay slightly higher fees without breaking user experience. The end of “free gas” might force the industry to grow up—to build sustainable business models rather than burning VC money on subsidized fees. That’s the hopeful pragmatist in me.
But let’s face the blind spots. The biggest one is that most rollups are not prepared. I’ve audited the code of over a dozen L2 projects, and I can tell you that few have built-in mechanisms for dynamic fee management. They simply assume blob space will always be cheap. When fees spike, they will have to scramble—potentially breaking their fee models, angering users, and losing market share. The second blind spot is regulatory: if blob fees rise, it could bring Ethereum’s L2 ecosystem under the scrutiny of antitrust regulators, who might see the base fee as a form of collusion. That’s a stretch, but not impossible.
I also worry about the “split brain” scenario. If alt-DA becomes too popular, Ethereum’s core rollup ecosystem could fragment. Users might follow the cheapest data availability, but that would compromise the security that makes Ethereum special. We’ve seen this before with the “scalability trilemma.” The market will choose convenience over security, and then cry when the security fails.
Takeaway: Surviving the Winter to Plant the Spring
So, what should you do? If you’re a rollup developer, start measuring your blob consumption today. Build in fee buffers. Experiment with alt-DA for non-critical data. If you’re a user, be prepared for higher fees in 2-3 years. Don’t assume the current low-cost environment is permanent. If you’re an investor, look for rollups that are efficient with data—those that use compression, those that have a clear plan for scaling, those that are not just “run on Ethereum” but “run smart on Ethereum.”
Surviving the winter to plant the spring. The bears will say blob saturation is another reason to be bearish on Ethereum. I see it differently. It’s a signal that the ecosystem is maturing. Scarcity forces innovation. The rollups that survive will be the ones that treat data as a precious resource, not a free lunch. And when the next bull market comes, those rollups will be the ones that thrive.
We don’t need more blind optimism. We need clear-eyed analysis. The blob ceiling is real, but it’s not a wall. It’s a speed bump. And if we navigate it well, the road ahead is still open. Code is law, but empathy is truth. The truth is, blob space is finite. Let’s use it wisely.
