Our discv5 PR just landed in ethlambda. Lean nodes can now find peers on their own instead of being handed a static bootnode list.
Off by default, you can enable it with --discovery.enable.
Last week we fixed a bug in our pruning logic that was quietly costing us CPU and time to finality. The shape of the graph tells the story: things get worse when pruning starts, and go back to normal once the fix lands.
ALT Graphs showing Finalization and Justification delay on the top, and CPU usage on the bottom. All graphs show an increase from the left baseline after pruning was enabled, and a decrease back to the baseline once the fix was deployed.
New blog post out: ethlambda, how we made our blocks arrive on time
In this post, we talk about how slow the proposer really is, and show how it can be made fast enough, without changing the underlying cryptographic primitives.
Proving time is unchanged, but the same work now lands comfortably inside the deadline.
Benchmarked on our servers (8 cores / 16 threads, 64 GB). The happy path here is 1-attestation blocks with a single proof per attestation (no compaction).
New blog post out: ethlambda, devnet 5 and beyond.
Each devnet introduced a new piece of the protocol's complexity. Devnet 5 is now live and in interop testing, and it's a good time to look back at how far we've come and what's on the roadmap ahead.
In case anyone missed it, yesterday started the EF devnet 5 client interop runs. We already merged our devnet5 branch to main and are running internal devnets in parallel, looking for issues or performance improvement opportunities.
After investigating, we found the cause: a recently-created 256 MB buffer was being copied into a new memory location owned by the new allocator used by the library.
We made a fix for this, which resulted in an overall 7% improvement from the old version.
We saw an aggregation performance regression in a devnet, when trying out the newest leanVM changes. When running leanVM benchmarks on it, and comparing both the old and new commits, we found that the regression affected the raw XMSS aggregation only (not recursive aggregation).
We're going live tomorrow on X with our Community Call at 3:30 PM UTC. We'll be sharing what the team has been working on and what's next on our roadmap.
Post-quantum aggregation is slow, but leanVM recently introduced performance improvements that made it much faster.
On Devnet 5, worst-case proposal latency dropped from 2.5s to 1.3s, nearly a 2x improvement over the previous version.
Full-block aggregation is coming to devnet 5.
We've been running initial tests on our servers (shown on the last community call). Here's a look at the dashboard.
Tomorrow at 3:30 PM UTC, we're going live on for our Community Call. Expect a full rundown of recent work, this week's highlights, and what's coming next. See you there.
Here's a preview of ethlambda driving @ethrex_client over the Engine API. First demo of an all-@class_lambda Ethereum stack!
State tuned for more updates.
Wondering where ethlambda stands?
We put together a progress page with everything: completed devnets, active work, upcoming milestones, and the full roadmap ahead.
This is what building in public looks like.
Tomorrow at 3:30 PM UTC, we're going live on X for our Community Call. Expect a full rundown of recent work, this week's highlights, and what's coming next. See you there.
Ethlambda is now passing 100% of the Hive lean-spec-tests suites, the official conformance tests that run Lean Consensus clients against the spec.
Support for the new suites was merged yesterday, and we're looking into the few failing ones from rpc-compat and reqresp.
We're going live tomorrow on X with our Community Call at 3:30 PM UTC.
We'll be sharing the latest updates, what the team has been working on, and what's next on our roadmap.
Join us to get the full overview.
Watching fork choice in action.
Ethlambda's forkchoice visualization just landed a fresh round of upgrades from the hands of our contributors.
With clearer block tags, cleaner weight visualization, and scrolling fixes, forks have never looked this good.
Today, ethlambda is passing all leanConsensus hive tests! (except for nlean interop; we're working on that). You can see that at a glance in this nice view recently added by @ReamLabs.
We improved signature aggregation performance, a key component for scaling post-quantum Ethereum signatures, achieving a 3× increase in throughput after a series of optimizations and careful profiling.
In this post, we cover the key bottlenecks and the changes made.
Join us tomorrow on X at 3:30 PM UTC for our Community Call.
This week we reached a 3x improvement in signature aggregation throughput. We’ll break down the changes, share the latest updates, and talk through what’s next.
New blog post!
After covering LMD-GHOST in our previous write-up, we now move to the second half of consensus: finality.
How does the protocol decide when a block is truly permanent and cannot be reorged?
This post walks through 3SF-mini as implemented by @ethlambda_lean.
We are hosting the #10 Ethlambda Community Call tomorrow and will walk through recent progress, key insights, and upcoming work.
The call will be streamed live on X at 3:30 PM UTC. Join us to be part of the conversation.
New blog post.
Continuing our series, we dive into how we implement LMD-GHOST within Lean Consensus, and what that means for chain selection and network behavior.
Link below.
Debugging consensus failures is difficult, especially without visualization tools. We recently added a live fork visualization for ethlambda (similar to forky).
Here's a quick demo.
1/3
See you tomorrow for the next ethlambda community call.
We’ll be live at 3:30 PM (UTC) to share updates, recent milestones, and what we’re building next.
Streaming on X.
A built-in metrics stack is coming to lean-quickstart!
We recently opened a PR adding Grafana+Prometheus+leanMetrics to lean-quickstart, making it even easier to run a local devnet with an observability stack.
New blog post!
Building a post-quantum Ethereum client with the help of shared tooling: what a Lean Consensus client consists of and what tooling facilitates its development.
Our third community call is happening this Friday at 3:30 PM (UTC).
If you’re following ethlambda or want to go deeper into what we’re building, this one’s for you.
We’re continuing the conversation.
Our second community call is this Friday at 3:30 PM (UTC). A space to share updates, ideas, and questions with the Ethlambda team.
Ethlambda community calls start this Friday.
These recurring sessions are open to everyone and focused on roadmap updates, technical context, and what’s coming next.
Join to listen, ask questions, or jump into the discussion.