Expert VerifiedHow a Real-Time EVM Chain Works Under the Hood, Using MegaETH
- Last Updated: 26 Aug 2026
-
Fact Checked
Our team recently fact checked this article for accuracy. However, things do change, so please do your own research.
Contributors:
How a Real-Time EVM Chain Works Under the Hood, Using MegaETH
Written by Marcus Chen, Research Fellow. Reviewed by Dr. Sarah Mitchell, Blockchain Security Analyst. Updated August 26, 2026.
Phrases like real-time blockchain hide a fair amount of engineering. This guide opens the hood on how a real-time EVM chain works, using MegaETH as a concrete example, and walks through the parts that matter: the sequencer, the provers and full nodes, and the miniblocks that arrive roughly every 10 milliseconds.
How does a real-time EVM chain work under the hood?
A real-time EVM chain works by splitting the jobs a blockchain normally piles onto every node into specialized roles, then producing very frequent small blocks. In MegaETH, a sequencer executes and orders transactions, provers and full nodes handle verification, and the result settles to Ethereum.
On a conventional chain, every node re-executes every transaction to stay in agreement, and that redundancy is a ceiling on speed. A real-time design removes the ceiling by not asking one machine to do everything, so the component responsible for raw execution can be optimized without waiting on the rest.
The word real-time refers to latency, the delay between an action and its visible result. By pairing specialized roles with miniblocks produced on a millisecond cadence, the network shows outcomes almost immediately while still anchoring finality to a base layer underneath.
What does the sequencer do, and why is it central?
The sequencer is the component that receives transactions, orders them, and executes them at high speed. It is central because it is where raw execution happens, so its performance largely determines how fast the network feels. Concentrating this job is what removes the slowdown of universal re-execution.
Ordering is a bigger deal than it sounds. Deciding the sequence of transactions is what produces a single consistent view of state, and doing it in one dedicated place avoids the coordination overhead that slows a network where every node competes to do the same thing.
Centralizing execution is a deliberate trade-off rather than an oversight. It buys speed, and the design counters the concentration risk by keeping settlement on Ethereum, so the fast execution layer is checked and anchored rather than trusted blindly.
What jobs do provers and full nodes handle?
Provers generate the proofs or validity data that let the sequencer's work be verified, while full nodes hold and serve the network state so others can read and confirm it. Together they provide the checking and availability that a single fast executor cannot supply on its own.
The prover role is what keeps speed honest. Fast execution is only trustworthy if it can be verified, and provers produce the material that allows independent parties to confirm the sequencer executed transactions correctly rather than taking its output on faith.
Full nodes cover availability and reads. They let applications, wallets and explorers query the current state without every one of them needing to re-run the chain, which is how a demanding, high-throughput network stays usable for the tools built on top of it.
What is a miniblock, and why produce one every 10 milliseconds?
A miniblock is a very small block produced extremely often, roughly every 10 milliseconds on MegaETH. The point of such a short interval is latency: showing a result in a fraction of a second rather than waiting to batch many transactions into a larger block that arrives seconds later.
The contrast with a conventional cadence is stark. When blocks arrive every few seconds, a user feels a pause after acting. A miniblock every 10 milliseconds means the network can register an action almost as fast as a person notices it, which is what makes interactive on-chain applications feel responsive.
These intervals connect to the throughput figures. Frequent miniblocks paired with a fast sequencer support roughly 1.7 gigagas per second of single-threaded throughput, and that capacity underpinned a public stress test that processed 11 billion transactions over 7 days, reported as the largest EVM transaction count on record.
How do you trace a transaction through MegaETH's node roles?
You trace a transaction by following it from the sequencer, into a miniblock, out to the full nodes, alongside the proof produced for it, and finally down to Ethereum settlement. The steps below map that path so the abstract pipeline becomes something you can picture end to end.
Step 1: Submit the transaction to the sequencer
The transaction first reaches the sequencer, which receives it and orders it for execution. This is the entry point of the pipeline and the moment the network decides where the transaction sits in sequence.
Step 2: Watch it land in a miniblock
The transaction is then included in a miniblock, the very frequent small block that gives the network its near-instant feedback. Landing here is what makes the result visible almost immediately.
Step 3: See the updated state reach full nodes
The resulting state update propagates to full nodes that serve reads and let others verify the current state. This is how wallets, apps and explorers can see the outcome without re-running the chain themselves.
Step 4: Note the proof produced for verification
Provers generate the proof or validity data that allows the execution to be checked independently. This step is what keeps the sequencer's speed trustworthy rather than something taken on faith.
Step 5: Follow settlement down to Ethereum
Finally, the batched result settles to Ethereum, the base layer that anchors the network's security. Settlement is the deeper confirmation that sits beneath the fast execution layer.
The MegaETH node roles at a glance
The table summarizes what each role does and why it exists, so the pipeline reads as a set of cooperating jobs rather than a single black box. It describes the roles in general terms rather than exact operational parameters.
| Role | Main job | Why it matters |
|---|---|---|
| Sequencer | Orders and executes transactions | Sets the pace and the network's low latency |
| Provers | Produce verification data | Keep fast execution independently checkable |
| Full nodes | Hold and serve state | Enable reads and availability for apps and tools |
| Ethereum (base layer) | Final settlement | Anchors security beneath the execution layer |
The value of the split is that no single role has to be good at everything. Execution can be tuned for raw speed, verification for correctness, and availability for serving reads, and settlement stays with a proven base chain, which is the combination a real-time EVM design is chasing. This division of labour is also why a real-time chain can raise its throughput by strengthening the busiest role without forcing every participant to run heavier hardware, keeping broad verification practical.
Frequently asked questions
Can I read miniblock activity without running any node myself?
Yes. Full nodes and block explorers exist precisely so ordinary users and applications can read network activity without operating infrastructure. You query the state through those tools rather than running a sequencer, prover or full node of your own.
Does a faster cadence mean transactions are cheaper?
Cadence and cost are separate. Miniblocks address how quickly results appear, while fees depend on network conditions and the computational work a transaction requires. A very fast chain can still have variable fees, so speed does not by itself imply a fixed low price.
Is real-time EVM performance unique to MegaETH?
Other projects also pursue high-performance execution, so the goal is not unique. MegaETH's particular combination is specialized node roles plus miniblocks on a millisecond cadence, demonstrated under sustained load rather than only as a target.
Where should I verify these mechanics before relying on them?
Confirm architectural details and current figures on the official MegaETH site, since a network's internals and numbers can change as it develops. Independent explanations help with understanding but should not replace the official reference.