Maximal Extractable Value (MEV) has been a prominent topic since its formalization in the Flashboys 2.0 paper in 2019. However, the question “what are the ways to decentralize block building?” only gained significant attention with the introduction of proposer-builder separation shortly after the Merge.
Recently, an exciting proposal that promises to solve the problem has been found, and we would like to talk about it — BuilderNet. But before that, let’s explore the current block building landscape on Ethereum and compare it to the impact that BuilderNet brings.
In classic Ethereum validator scenario, transactions are picked up from the mempool and ordered in a block based on priority fees. The block is then proposed for addition to the canonical chain, which involves signing it, gossiping it to peers, and receiving attestations.
Proposer-builder separation (PBS) divides the block creation process by allowing block builders to create and offer blocks to proposers, who select the most profitable option without seeing its contents. This separation is crucial because building is a specialized task requiring high-performance software and hardware, and having validators perform this task could harm decentralization. Builders often run algorithms to maximize value extraction from each block, while proposers pay a fee to the builders before broadcasting the selected block.
Separating block builders and proposers not only makes transaction censorship more difficult, but also helps to ensure broader participation in MEV extraction. Without a builder marketplace, only validators with the necessary hardware could participate, leaving smaller stakers and solo validators at a disadvantage. PBS also counters the centralizing effects of MEV, where large validators could exploit economies of scale, reinvesting profits into better hardware and proprietary algorithms to maximize MEV revenue.
This would ultimately lead to centralization, as smaller validators either become unprofitable or are overshadowed by larger ones. Additionally, by giving proposers the power to constrain block builders through inclusion lists, PBS reduces the risk of censorship, ensuring that block builders can’t unilaterally decide which transactions make it into blocks.
Flashbots introduced MEV-Boost, an open-source middleware enabling validators to tap into a competitive block-building market. It is the most popular implementation of PBS for Ethereum. MEV-Boost created an open market for block builders, enhancing competition and censorship resistance on Ethereum.
Source: https://buildernet.org/blog/introducing-buildernet
At first glance, MEV-Boost, developed by Flashbots, seems like a perfect solution for block building on Ethereum. However, it has several downsides and areas for improvement.
Currently, over 90% of Ethereum blocks are built by just two entities: Beaverbuild and Titan Builder. This significant concentration, which has tripled since 2022, is a widely recognized issue.
Source: https://mevboost.pics/
Exclusive off-chain deals between orderflow providers and builders have entrenched a duopoly, undermining Ethereum’s neutrality, censorship resistance, and resilience. Block building faces a significant risk of full centralization, especially in the absence of inclusion lists that enable proposers to force-include specific transactions, effectively granting a few builders control over block contents. Additionally, block building follows a power-law distribution, making monopolies difficult to disrupt and further centralization inevitable without proactive measures to decentralize block production.
To counter this, Flashbots and collaborators have spent two years exploring ways to decentralize block building, including programmable privacy (MEV-Share), distributed block building (MEVM), and secure hardware solutions (Trusted Execution Environment). Since Ethereum improves censorship resistance through protocol changes like inclusion lists, it must also decentralize block building to strengthen its defenses.
Ethereum promises a resilient, permissionless platform for global finance and innovation. Adding new blocks to this ledger is best handled by sophisticated block builders who optimize complex needs across users and apps. By distributing block building across a broad network, we can prevent chokepoints, foster global coordination, and create more value for users, apps, and wallets.
BuilderNet enables collaborative block building, offering an open, neutral alternative to costly and inefficient orderflow competition. By simplifying participation, it allows builders, searchers, and apps to focus on efficiency, innovation, and user value instead of worrying about censorship resistance and MEV leakage.
Apps and wallets aim to internalize MEV and offer users affordable, simple transactions. BuilderNet’s first release uses an open-source refund rule to compensate users for their contributions to blocks. Future upgrades will refine this rule to support diverse use cases, make refunds fully permissionless, and fairly compensate all network participants, including TEE operators.
Currently, landing a bundle on Ethereum requires trusting a third party or running your own block builder. BuilderNet changes this by using TEEs to ensure privacy and integrity, while providing refunds to searchers who contribute to its blocks. This creates an even playing field, giving independent searchers the same financial outcomes and privacy as integrated searcher-builders.
For validators, BuilderNet won’t initially affect their economics, as block builders already withhold value from validators. However, over time, BuilderNet aims to shift the market, allowing orderflow providers to retain more MEV and creating a simpler, more transparent gas market.
Flashbots has already retired its centralized block builders and shifted the orderflow and refunds to BuilderNet. Beaverbuild will do the same in the future, merging their software, infrastructure, and orderflow with the decentralized network. Beaverbuild is starting by running builder instances in TEEs, allowing gradual adoption of decentralization with minimal disruption for users. This will serve as a model for integrating other block builders with BuilderNet.
Source: https://www.relayscan.io
The first release of BuilderNet introduces a multioperator system, enabling multiple parties to operate the same block builder. Each operator runs an instance of an open-source builder in a TEE, allowing orderflow providers to verify and send encrypted orderflow.
The primary reason BuilderNet uses TEEs is to enable multiple parties to operate block builders while maintaining privacy and integrity. TEEs ensure that encrypted data is not leaked to operators and provide integrity guarantees through remote attestations, allowing orderflow providers to verify correct software use. New TEEs like Intel TDX offer these benefits without sacrificing speed. However, TEEs come with certain downsides, including performance overhead due to security measures, reliance on specific hardware, and vulnerabilities to TEE-targeted exploits.
All BuilderNet block builder instances currently run in TDX, and the system employs a verifiable boot chain and integrity-protected runtime environment, ensuring reproducible verification of VM images and attested instances.
Each instance shares the received orderflow with others in the network and submits blocks to MEV-Boost relays as usual. When a BuilderNet instance wins a block, refunds are calculated and distributed to orderflow providers based on the value they contributed.
Source: https://buildernet.org/docs/architecture
BuilderNet’s refund mechanism shares value with those who contribute to its blocks, aiming to provide an efficient, transparent way for orderflow providers to internalize MEV and support the network’s operations.
Users, wallets, apps, and searchers who submit transactions or bundles to BuilderNet will receive refunds based on their contribution to its blocks. Refunds are calculated using an open-source rule, which evaluates the value each transaction adds to the block by simulating the block with and without it. The remainder is shared proportionally among users based on their contribution. Transactions sent to Flashbots Protect, Flashbots Bundle Relay, or directly to a BuilderNet node will automatically receive refunds, but public mempool transactions will not. In the future, the refund rule will also compensate operators for their computation costs.
The first release of BuilderNet is a step toward decentralizing block building on Ethereum. It introduces key guarantees today and outlines plans for future upgrades.
In the future, each BuilderNet operator will be able to set their own compliance policy, allowing them to exclude certain transactions from their blocks. This enhances censorship resistance, as different operators may include transactions that others exclude. Over time, BuilderNet plans to enable operators to extend each other’s blocks, ensuring that even a single non-censoring operator can prevent censorship. This will make BuilderNet more resistant to censorship than the current MEV-Boost market.
Currently, all block builders are operated by centralized teams. BuilderNet enables multiple teams to operate instances of the same builder, increasing decentralization. At launch, all orderflow sent to Flashbots will also be shared with BuilderNet, broadening the number of teams involved. However, BuilderNet is not fully decentralized yet; it still relies on centralized infrastructure and requires permission to join. Future releases will make all roles permissionless and decentralize governance.
Ultimately, BuilderNet will create more value than centralized builders by enabling operators to outsource computations to private searchers and bundle-merging programs, increasing block value and MEV capture.
BuilderNet represents a significant step toward decentralizing block building on Ethereum, addressing the key challenges of MEV and censorship resistance. While it is still in its early stages, this approach offers a promising path towards decentralized blockchain infrastructure. As more builders and orderflow providers join the network, BuilderNet will continue to evolve, ensuring a fairer distribution of MEV and a stronger, more sustainable Ethereum Blockchain.
Maximal Extractable Value (MEV) has been a prominent topic since its formalization in the Flashboys 2.0 paper in 2019. However, the question “what are the ways to decentralize block building?” only gained significant attention with the introduction of proposer-builder separation shortly after the Merge.
Recently, an exciting proposal that promises to solve the problem has been found, and we would like to talk about it — BuilderNet. But before that, let’s explore the current block building landscape on Ethereum and compare it to the impact that BuilderNet brings.
In classic Ethereum validator scenario, transactions are picked up from the mempool and ordered in a block based on priority fees. The block is then proposed for addition to the canonical chain, which involves signing it, gossiping it to peers, and receiving attestations.
Proposer-builder separation (PBS) divides the block creation process by allowing block builders to create and offer blocks to proposers, who select the most profitable option without seeing its contents. This separation is crucial because building is a specialized task requiring high-performance software and hardware, and having validators perform this task could harm decentralization. Builders often run algorithms to maximize value extraction from each block, while proposers pay a fee to the builders before broadcasting the selected block.
Separating block builders and proposers not only makes transaction censorship more difficult, but also helps to ensure broader participation in MEV extraction. Without a builder marketplace, only validators with the necessary hardware could participate, leaving smaller stakers and solo validators at a disadvantage. PBS also counters the centralizing effects of MEV, where large validators could exploit economies of scale, reinvesting profits into better hardware and proprietary algorithms to maximize MEV revenue.
This would ultimately lead to centralization, as smaller validators either become unprofitable or are overshadowed by larger ones. Additionally, by giving proposers the power to constrain block builders through inclusion lists, PBS reduces the risk of censorship, ensuring that block builders can’t unilaterally decide which transactions make it into blocks.
Flashbots introduced MEV-Boost, an open-source middleware enabling validators to tap into a competitive block-building market. It is the most popular implementation of PBS for Ethereum. MEV-Boost created an open market for block builders, enhancing competition and censorship resistance on Ethereum.
Source: https://buildernet.org/blog/introducing-buildernet
At first glance, MEV-Boost, developed by Flashbots, seems like a perfect solution for block building on Ethereum. However, it has several downsides and areas for improvement.
Currently, over 90% of Ethereum blocks are built by just two entities: Beaverbuild and Titan Builder. This significant concentration, which has tripled since 2022, is a widely recognized issue.
Source: https://mevboost.pics/
Exclusive off-chain deals between orderflow providers and builders have entrenched a duopoly, undermining Ethereum’s neutrality, censorship resistance, and resilience. Block building faces a significant risk of full centralization, especially in the absence of inclusion lists that enable proposers to force-include specific transactions, effectively granting a few builders control over block contents. Additionally, block building follows a power-law distribution, making monopolies difficult to disrupt and further centralization inevitable without proactive measures to decentralize block production.
To counter this, Flashbots and collaborators have spent two years exploring ways to decentralize block building, including programmable privacy (MEV-Share), distributed block building (MEVM), and secure hardware solutions (Trusted Execution Environment). Since Ethereum improves censorship resistance through protocol changes like inclusion lists, it must also decentralize block building to strengthen its defenses.
Ethereum promises a resilient, permissionless platform for global finance and innovation. Adding new blocks to this ledger is best handled by sophisticated block builders who optimize complex needs across users and apps. By distributing block building across a broad network, we can prevent chokepoints, foster global coordination, and create more value for users, apps, and wallets.
BuilderNet enables collaborative block building, offering an open, neutral alternative to costly and inefficient orderflow competition. By simplifying participation, it allows builders, searchers, and apps to focus on efficiency, innovation, and user value instead of worrying about censorship resistance and MEV leakage.
Apps and wallets aim to internalize MEV and offer users affordable, simple transactions. BuilderNet’s first release uses an open-source refund rule to compensate users for their contributions to blocks. Future upgrades will refine this rule to support diverse use cases, make refunds fully permissionless, and fairly compensate all network participants, including TEE operators.
Currently, landing a bundle on Ethereum requires trusting a third party or running your own block builder. BuilderNet changes this by using TEEs to ensure privacy and integrity, while providing refunds to searchers who contribute to its blocks. This creates an even playing field, giving independent searchers the same financial outcomes and privacy as integrated searcher-builders.
For validators, BuilderNet won’t initially affect their economics, as block builders already withhold value from validators. However, over time, BuilderNet aims to shift the market, allowing orderflow providers to retain more MEV and creating a simpler, more transparent gas market.
Flashbots has already retired its centralized block builders and shifted the orderflow and refunds to BuilderNet. Beaverbuild will do the same in the future, merging their software, infrastructure, and orderflow with the decentralized network. Beaverbuild is starting by running builder instances in TEEs, allowing gradual adoption of decentralization with minimal disruption for users. This will serve as a model for integrating other block builders with BuilderNet.
Source: https://www.relayscan.io
The first release of BuilderNet introduces a multioperator system, enabling multiple parties to operate the same block builder. Each operator runs an instance of an open-source builder in a TEE, allowing orderflow providers to verify and send encrypted orderflow.
The primary reason BuilderNet uses TEEs is to enable multiple parties to operate block builders while maintaining privacy and integrity. TEEs ensure that encrypted data is not leaked to operators and provide integrity guarantees through remote attestations, allowing orderflow providers to verify correct software use. New TEEs like Intel TDX offer these benefits without sacrificing speed. However, TEEs come with certain downsides, including performance overhead due to security measures, reliance on specific hardware, and vulnerabilities to TEE-targeted exploits.
All BuilderNet block builder instances currently run in TDX, and the system employs a verifiable boot chain and integrity-protected runtime environment, ensuring reproducible verification of VM images and attested instances.
Each instance shares the received orderflow with others in the network and submits blocks to MEV-Boost relays as usual. When a BuilderNet instance wins a block, refunds are calculated and distributed to orderflow providers based on the value they contributed.
Source: https://buildernet.org/docs/architecture
BuilderNet’s refund mechanism shares value with those who contribute to its blocks, aiming to provide an efficient, transparent way for orderflow providers to internalize MEV and support the network’s operations.
Users, wallets, apps, and searchers who submit transactions or bundles to BuilderNet will receive refunds based on their contribution to its blocks. Refunds are calculated using an open-source rule, which evaluates the value each transaction adds to the block by simulating the block with and without it. The remainder is shared proportionally among users based on their contribution. Transactions sent to Flashbots Protect, Flashbots Bundle Relay, or directly to a BuilderNet node will automatically receive refunds, but public mempool transactions will not. In the future, the refund rule will also compensate operators for their computation costs.
The first release of BuilderNet is a step toward decentralizing block building on Ethereum. It introduces key guarantees today and outlines plans for future upgrades.
In the future, each BuilderNet operator will be able to set their own compliance policy, allowing them to exclude certain transactions from their blocks. This enhances censorship resistance, as different operators may include transactions that others exclude. Over time, BuilderNet plans to enable operators to extend each other’s blocks, ensuring that even a single non-censoring operator can prevent censorship. This will make BuilderNet more resistant to censorship than the current MEV-Boost market.
Currently, all block builders are operated by centralized teams. BuilderNet enables multiple teams to operate instances of the same builder, increasing decentralization. At launch, all orderflow sent to Flashbots will also be shared with BuilderNet, broadening the number of teams involved. However, BuilderNet is not fully decentralized yet; it still relies on centralized infrastructure and requires permission to join. Future releases will make all roles permissionless and decentralize governance.
Ultimately, BuilderNet will create more value than centralized builders by enabling operators to outsource computations to private searchers and bundle-merging programs, increasing block value and MEV capture.
BuilderNet represents a significant step toward decentralizing block building on Ethereum, addressing the key challenges of MEV and censorship resistance. While it is still in its early stages, this approach offers a promising path towards decentralized blockchain infrastructure. As more builders and orderflow providers join the network, BuilderNet will continue to evolve, ensuring a fairer distribution of MEV and a stronger, more sustainable Ethereum Blockchain.