Today, the Glamsterdam upgrade, which is second only to The Merge in importance for Ethereum, will officially complete its testnet activation, and Ethereum will officially enter the era of parallel blockchain with one foot. Ethereum has undergone 20 upgrades in its history, the most important of which was The Merge in 2022, which completely shifted Ethereum from POW to POS and changed the economic model of Ethereum. Of course, everyone is more familiar with the Dencun upgrade through L2 expansion in 2024, but with the chaos of the Nine Sons of L2 and Vitalik personally overturning this route, the history of the Zhou Dynasty's enfeoffment of the world's princes was ended. To build a unified empire of the Qin Dynasty, the expansion of the Ethereum main network became a top priority, and this time Glamsterdam is equivalent to replacing Ethereum's engine. The core reason why the concept of parallel blockchain was hailed as the killer of Ethereum in recent years is that the vast majority of blockchains, including Ethereum, can only execute transactions in sequence A → B → C → D. Therefore, in order to improve performance, the number of nodes can only be reduced at the expense of decentralization. The reason why ABCD cannot run simultaneously is that there may already be some dependency relationships between Ethereum transactions, and nodes do not know whether several transactions can be executed simultaneously, so they can only execute them one by one in a very cumbersome and primitive way. This upgrade introduces EIP-7928 called Block Level Access Lists, abbreviated as BAL. It records in advance which accounts, storage spaces, and final states each block has accessed, so that client nodes can know in advance whether there are dependencies between each transaction. Therefore, they can execute transactions that do not have dependencies at the same time. According to historical data backtesting, about 70% of Ethereum transactions do not have dependencies. But BAL also has a side effect of generating additional costs for data propagation and storage, so this Glamsterdam is a combination of punches, in addition to BAL, it also adds ePBS. EIP-7732 is called Enshrined Proposer Builder Separation, abbreviated as ePBS, which solves the historical legacy problem of who is responsible for proposing blocks and who is responsible for building blocks. The Proposer proposer is responsible for deciding which validator is responsible for issuing blocks today, while the Builder builder is responsible for selecting the most profitable group from thousands of transactions and putting it into the block. However, there is no regulation on how the builder should quote the proposer during this process, so the industry has created a bunch of roles, including MEVs, intermediaries, and other "small vendors" to coordinate. As a result, the birth of a block now goes through validators → MEVs → intermediaries → builders → execution verification, and a large number of relationships do not belong to the Ethereum consensus layer itself. In the core block generation process of Ethereum, there are actually a lot of things outside the protocol inserted, which requires proposers to complete propagation, execution, and verification within a very short time window after receiving a block. If all of these can be stuffed into Ethereum's consensus layer, this time window can be extended without being so tight. EPBS requires the builder to commit in advance to preparing the resources for the block, and then provide confirmation to the proposer. The two parties reach an agreement to ensure that everything that needs to be prepared for this block is prepared in advance, so that the remaining propagation, execution, and verification processes of the block do not need to be rushed. So ePBS can expand the blocks, thereby solving to some extent the problem of data growth caused by BAL mentioned earlier. In short, the Glamsterdam upgrade of Ethereum has further enhanced the speed and performance of the mainnet, and the historical mission of L2 scalability will be completely terminated.