The recent "Glamsterdam" test on the Ethereum network has demonstrated a remarkable increase in the amount of gas that can be processed within a single block, approaching the 200 million gas mark. This milestone follows a series of upgrades and optimizations that were introduced to the Sepolia testnet, which serves as a proving ground for new protocol changes before they are deployed on the main Ethereum network. ### Background on the Glamsterdam Test The term "Glamsterdam" is a playful nod to the famed Dutch city known for its vibrant culture, and it has been adopted by the Ethereum community to describe a series of high‑throughput experiments aimed at pushing the limits of the network’s capacity. These experiments are conducted on Sepolia, a public testnet that mirrors the mainnet’s consensus rules but allows developers to experiment without risking real assets.
The primary goal of the Glamsterdam series is to evaluate how well recent protocol upgrades—particularly those targeting scalability and efficiency—perform under heavy load. ### The Upgrade and Its Impact The latest upgrade, which was rolled out in the weeks leading up to the Glamsterdam trial, introduced several key changes: 1. **EIP‑4844 (Proto‑Danksharding) Integration** – This proposal adds a new transaction type that carries a "blob" of data separate from the main transaction payload.
By off‑loading large data to these blobs, the core execution layer can handle more transactions per block. 2. **Optimized Gas Accounting** – Adjustments to how gas is charged for certain opcodes reduce the overall gas cost of common operations, freeing up capacity for additional transactions.
3. **Improved P2P Propagation** – Enhancements to the network’s peer‑to‑peer layer reduce latency and improve the speed at which new blocks are disseminated, which is crucial when blocks become densely packed with transactions. Collectively, these upgrades were expected to increase the block gas limit, but the actual observed performance during the Glamsterdam test surpassed many internal forecasts. ### Processing Budget Expansion Prior to the upgrade, Sepolia’s processing budget—a metric that defines the maximum amount of computational work that can be performed in a block—was set at roughly 70 million gas.
After the upgrade, the budget was increased more than threefold, reaching an allocation of about 220 million gas per block. This dramatic rise was intentional: developers wanted to see how the network behaved when the theoretical ceiling was pushed far beyond typical usage patterns. Despite this generous allocation, the sampled blocks during the test period consistently used less than half of the available gas.
On average, blocks contained between 90 million and 110 million gas worth of transactions. This under‑utilization suggests that, even with the new capacity, the current demand on Sepolia does not yet saturate the network.
It also indicates that the upgrades have successfully mitigated many of the bottlenecks that previously limited block utilization. ### Why the Gap Between Budget and Usage? Several factors contribute to the observed disparity between the allotted processing budget and actual gas consumption: - **Transaction Volume** – Sepolia, being a testnet, naturally sees fewer transactions than the mainnet.
Most developers run controlled experiments rather than the continuous stream of user‑generated activity seen on Ethereum’s live network. - **Gas‑Heavy Transactions** – The introduction of blob‑carrying transactions (EIP‑4844) means that a single transaction can now carry a large amount of data without consuming proportional gas. This changes the dynamics of gas usage, allowing more logical operations to fit within the same gas budget. - **Network Health Safeguards** – Clients implement safety checks that may deliberately limit the number of transactions packed into a block to avoid over‑loading nodes, especially during experimental phases.
### Implications for the Mainnet The success of the Glamsterdam test carries significant implications for the Ethereum mainnet, especially as the network moves toward full implementation of Danksharding and other scalability solutions. If a testnet can comfortably handle close to 200 million gas per block while only utilizing half of its allotted budget, the mainnet—once it adopts similar upgrades—could see a substantial increase in transaction throughput without compromising security or decentralization.
Moreover, the test provides valuable data for developers building roll‑up solutions and other layer‑2 protocols. Knowing that the underlying consensus layer can accommodate larger batches of data means that roll‑ups can be designed to submit more aggregated transactions, reducing overall fees and improving user experience.
### Future Directions The Ethereum research community plans to continue the Glamsterdam series with even more aggressive parameters. Upcoming experiments will: - **Increase the Blob Size** – Testing larger data blobs to see how they affect block propagation and verification times.
- **Stress‑Test Node Implementations** – Running the same high‑throughput scenario across multiple client implementations (e.g., Geth, Nethermind, Besu) to ensure interoperability. - **Simulate Real‑World Load** – Introducing synthetic traffic that mimics the transaction mix of the mainnet, including DeFi swaps, NFT mints, and cross‑chain bridges. These efforts aim to fine‑tune the balance between block size, latency, and security, ensuring that when the upgrades finally roll out to the live network, they do so with confidence. ### Conclusion The Glamsterdam test on Sepolia has shown that after a comprehensive upgrade, the network can process nearly 200 million gas per block—a figure that dwarfs the previous limits and demonstrates the effectiveness of recent protocol enhancements.
While the actual gas usage during the test remains well below the expanded budget, this headroom is a positive sign that Ethereum’s scalability roadmap is on track. As the community moves forward, the insights gained from these experiments will be instrumental in shaping a faster, more efficient, and more resilient Ethereum mainnet for the years to come.