I remember the quiet before the Dencun upgrade. Sitting in a coworking space in Manila, watching the countdown ticker on Etherscan, I felt a familiar unease. It was the same feeling I had in 2017 when I analyzed 40+ whitepapers during the ICO boom — the sense that everyone was celebrating a solution before the problem had fully revealed itself. The Dencun upgrade introduced blobs, a temporary data layer designed to make rollups cheaper. The market cheered. Gas fees for L2s dropped by 90% overnight. But I spent the next three months digging into the math, and what I found kept me awake at night: blob data will be saturated within two years, and then all rollup gas fees will double again.
We burned out trying to own the future. But the future of scalability is not about owning more space — it's about understanding the limits of the space we have already built.
Context: The Historical Narrative Cycles of Data Availability
To understand the blob saturation problem, we need to walk back through the narrative cycles of Ethereum scalability. In 2020, during the DeFi Summer, I interviewed twelve early adopters for my series "The Illusion of Decentralized Wealth." Back then, the bottleneck was execution — Ethereum could barely handle 15 TPS. Rollups promised to fix that by moving execution off-chain. By 2023, rollups were live, but the bottleneck shifted to data availability: posting transaction data to L1 was costing millions. The industry consensus was "we need more data space." Enter Dencun and EIP-4844 — blobs.
Blobs are essentially temporary data packets that are stored by beacon nodes for about 18 days. They are not executed by the EVM, so they provide cheap bandwidth for rollups. The narrative was simple: blobs would make L2 fees negligible, and Ethereum would scale infinitely. But this narrative ignores the fundamental constraint of supply and demand. Currently, Ethereum targets 3 blobs per block, with a maximum of 6. Each blob is ~128 KB. That gives us roughly 384 KB per block (at 3 blobs) or 768 KB (at 6). Blocks are produced every 12 seconds. So the theoretical maximum blob data per day is about 5.5 GB (at 6 blobs/block). But the average is closer to 2.7 GB/day because the network rarely hits the max.
Now, let's look at the demand side. According to data from Dune Analytics and my own analysis of on-chain blob usage, the top L2s — Arbitrum, Optimism, Base, and zkSync — are already consuming about 60% of the available blob space. In the last 30 days, blob usage has grown by 35% month-over-month. If this growth rate continues (and it will, as more L2s launch and existing ones increase throughput), we will hit the 3-blob target ceiling within 12 months. Once that happens, the fee market for blobs will kick in. Base fees will rise, and rollup operators will bid for space. The result: L2 gas fees will increase by a factor of 2x to 5x, depending on demand.
Core: The Narrative Mechanism of Saturation and Sentiment Analysis
Let me walk you through the technical mechanics. Blobs are priced via a separate fee market called the "blob gas market." It uses the same EIP-1559 algorithm as the execution layer, but with a target of 3 blobs per block. When the number of blobs in a block exceeds 3, the base fee increases. When it drops below, the base fee decreases. Currently, the base fee is often zero because we are under the target. But as more L2s start posting blobs, we will hit the target consistently. Once the demand exceeds 3 blobs per block, the base fee will rise exponentially.
I analyzed the blob usage patterns over the past month using a custom script that pulls data from the Beacon API. Here is what I found: Base (the Coinbase L2) is the largest blob consumer, followed by Arbitrum and Optimism. Base alone accounts for 35% of all blob transactions. The average blob size is around 120 KB, but some L2s are using compression to pack more data into each blob. This means that the effective data throughput is increasing even without more blobs. But compression has limits — eventually, the L2s will need to post more blobs to support their user growth.
Now, consider the sentiment. The market is still euphoric about Dencun. Every week, a new L2 announces its launch. The total number of rollups has grown from 10 in 2023 to over 40 today. Each one needs blob space. The narrative is that "Ethereum is becoming a settlement layer for rollups," and that is true. But the settlement layer has a fixed capacity. The blobs are not a permanent solution — they are a stopgap until sharding (or danksharding) arrives, which is likely years away.
Based on my audit experience from the ICO era, I've learned to be skeptical of solutions that promise to solve a problem before the problem is fully understood. The blob architecture was designed with the assumption that rollups would not need more than 3 blobs per block for a long time. But the adoption curve is steeper than anyone anticipated. In my conversations with L2 developers, they admit that they are already planning for higher blob usage. Some are even considering moving to alternative data availability layers like Celestia or EigenDA to avoid Ethereum's blob fees. That would fragment the ecosystem and reduce the value of Ethereum's security.
Contrarian Angle: The Blind Spot of Blob Consumption
Here is the contrarian take that almost no one is talking about: the blob saturation may actually be a feature, not a bug. High blob fees will force L2s to become more efficient. They will have to optimize their data compression, reduce unnecessary data, and possibly even batch transactions more aggressively. This could lead to a healthier ecosystem where only the most efficient rollups survive. It is similar to how high gas prices on Ethereum L1 forced developers to optimize smart contracts. The same Darwinian pressure will apply to L2s.
But there is a darker blind spot. The current blob fee market does not account for the cost of long-term storage. Blobs are only stored for 18 days. After that, they are pruned. This means that historical rollup data is not available on Ethereum — it must be stored elsewhere. If an L2 ceases to exist, or if its off-chain data is lost, the funds on that L2 could become unrecoverable. This is a systemic risk that the market is ignoring. The narrative of "Ethereum as the ultimate data availability layer" is incomplete because blobs are temporary. The real data availability is off-chain.
From my perspective, this is a replay of the 2017 ICO mania, where projects promised decentralized storage but actually stored data on centralized servers. The same pattern is emerging: L2s claim to use Ethereum for data availability, but the actual data is only available for 18 days. After that, it's effectively gone from the chain. If you need to prove a transaction on an L2 that happened two years ago, you will have to rely on the L2's own infrastructure or a third-party archiver. That is not decentralization.
Takeaway: The Next Narrative Shift
The next narrative shift will be from "blob space is cheap" to "blob space is a premium resource." The market will start pricing in the cost of data availability. We will see L2s compete for blob space, and the most efficient ones will win. The ones that rely on expensive data compression or frequent blobs will be priced out. This is already happening: some L2s are now offering "zero-fee" transactions to attract users, only to subsidize the blob costs through their native tokens. This is unsustainable in a bear market.
What does this mean for the average user? If you are using an L2 today, you should be aware that the fees you are paying are artificially low. They will rise. The only question is when. I predict that within 18 months, the average L2 transaction fee will be at least $0.10, and possibly higher for complex operations. The era of sub-cent fees is ending.
As I wrote in "The Silence After the Storm" in 2023, resilience is about understanding the limits of the systems we trust. The blob space is a limit. We have to plan for it. The next generation of L2s will need to be data-efficient, or they will fail. The market will punish those that waste blob space.
Final Thought
We burned out trying to own the future. But the future of Ethereum scalability is not about owning more blobs — it's about using the ones we have wisely. The narrative is shifting from abundance to scarcity. Are you ready for that shift?
