Vitalik Buterin supports minimum consensus layer for Ethereum over Casper FFG

TLDR

Minimit reached the final stage in one signature round, replacing Casper FFG's two-round validation and finalization process. (pieces to assemble – 105 possibilities)

The new gadget lowers the margin of error from 33 percent to 17 percent, but raises the unilateral censorship rate from 67 percent to 83 percent.

Buterin argued that since censorship is fast and lacks on-chain evidence, it poses a greater risk than eventual reversal.

Minimmit requires 83% of customers to report a bug before bug completion, giving developers a wider margin of safety.

Minimmit is proposed as a direct replacement for Casper FFG in the Ethereum consensus layer. Ethereum founder Vitalik Buterin recently shared a detailed technical post comparing both completion widgets.

Casper FFG has long been used as a two-round completion method on the network. The proposed system, on the contrary, reaches the final stage in one round of verification signatures.

The idea is gaining attention as the Ethereum community continues to evaluate changes to the consensus.

Why does the new system work in one round?

Casper FFG requires each validator to sign a block on two separate occasions. The first signature “confirms” the block, and the second “completes”.

Minimit reduces this to one signature round. This makes the process more efficient for authenticators in the network.

The change comes at a direct cost to fault tolerance, though. The new system's threshold is set at 17%, compared to 33% under the Casper FFG.

A small amount of malicious share can spoil the final stage of the new model. Still, Buterin's post makes the case that other features of the system more than make up for this drop.

In a post shared on X, Buterin described himself as a long-time “security speculation hawk” in Ethereum consensus research. Under synchronization it mentions a push past 49% fault tolerance.

He also cited his work on DAS for honesty-most-defensive data availability checks. Despite this record, he stated that he was “even enthusiastic” about the proposed design.

The asynchronous networking issue also differs between the two systems. Under a good 3SF, as long as the attacker controls less than 33% of the share, it remains final.

The proposed gadget reduces that same protection to 17%. In both cases, any overturning of finality will result in a large penalty on the guilty verifiers.

Censorship resistance and a broader security picture

Buterin's argument focuses on identifying censorship as a dangerous threat. Unlike the last copy, censorship does not immediately provide publicly verified evidence against the attacker.

A reverse event, on the other hand, results in automatic, large-scale disruption. This asymmetry is the main reason behind the support for the minimit design.

Both systems require an attacker to control more than 50% of ETH to censor. The main difference is in what happens at higher levels.

With 3SF, over 67% of attackers can single-handedly complete the chain, eliminating any coordination points for trusted validators. The new system raises this level to 83%.

Software bugs provide another area where the proposed gadget can take advantage. Under 3SF, a defect shared by 67% of client software could accidentally end up in a faulty chain state.

Minimit raises that level to 83%. This wider margin gives developers more time to identify and respond to bugs before they become permanent.

Buterin also cited an economic argument against the latest shift attacks. With a stake of 15 million ETH, it would require a reduction of 5 million ETH, or about $10 billion, to restore the final level of 3SF.

He pointed out that the starting point of 17 percent still represents a significant obstacle in itself. From there, other properties of the proposed system make it a stronger overall consensus for Ethereum.