Ethereum’s Medalla Testnet Launched With Five Clients—and a Finality Gap

Ethereum’s Medalla public multi-client testnet went live at 13:00:08 UTC on August 4, 2020, creating the largest community-run rehearsal yet for the planned first phase of Ethereum’s proof-of-stake system. The network began producing blocks, but validator participation started below the threshold needed for prompt finality.

The launch mattered because Medalla was not another centrally coordinated developer network. Ethereum Foundation coordinator Danny Ryan described it before genesis as a community-operated dress rehearsal: most validators would be run by outside participants using independently developed software. A stable run would put the planned proof-of-stake mainnet next in the development sequence, although no production launch occurred on August 4 and no mainnet date was guaranteed.

A fixed genesis with five clients

The Ethereum Foundation’s August 3 record set genesis at Unix time 1596546008—eight seconds after 13:00 UTC on August 4—after the required minimum of more than 16,000 validator deposits had been met. The configured threshold was 16,384 validators. At 32 units of test ETH per validator, that minimum represented 524,288 test ETH; this is a protocol calculation, not a measurement of unique people or capital.

Five consensus clients were expected at genesis: Teku, Prysm, Nimbus, Lodestar and Lighthouse. Their coexistence was central to the test. Each implementation had to interpret the same Phase 0 specification, exchange blocks and attestations, and agree on the chain without one client team coordinating every operator.

CoinDesk’s event-day report, citing the Beaconcha.in explorer, recorded more than 20,000 validators and about 650,000 test ETH associated with the network shortly after launch. Those figures were a point-in-time explorer snapshot, not an audit of distinct operators. One person or service could control many validator keys, and Medalla’s test ETH was not production ether and carried no claim on real ETH.

Blocks arrived before finality

Medalla produced blocks at its scheduled genesis, but the opening was uneven. Prysmatic Labs’ August 7 account put active participation just below 57% during the first half-hour. The client team said the chain required at least 66% participation to finalize, so Medalla initially advanced without finalizing checkpoints.

That distinction is important. Block production showed that the network had started; finality required a supermajority of validator stake to attest consistently enough to make checkpoints economically settled under the test protocol. The shortfall did not mean Ethereum’s existing proof-of-work mainnet had stopped, and it put no production ETH at risk.

The available evidence pointed to an operational coordination problem rather than a single demonstrated protocol failure. Prysmatic Labs attributed the weak opening to validators using outdated or misconfigured software and to operators coming online late amid last-minute client updates. On August 4, the complete distribution of causes was not independently established.

Why the rehearsal mattered

Medalla tested more than consensus code. It tested documentation, key generation, the validator Launchpad, operator readiness, peer-to-peer connectivity and client interoperability under a fixed public start time. That made the rough opening useful evidence: a network can satisfy its deposit condition while still lacking enough correctly configured, online validators at genesis.

No cryptocurrency price, return or trading-volume move is attributed to Medalla here. Crypto traded continuously across venues, and the reviewed sources do not establish that the testnet caused a measurable market move on August 4.

Later clarification

In its August 7 update, Prysmatic Labs said participation rose to roughly 80%–85% after more operators came online and updated nodes, allowing the chain to finalize. That later account clarifies the opening hours; it was not fully knowable at genesis. Medalla’s August 4 record therefore supports a narrower conclusion: Ethereum achieved a public, five-client proof-of-stake testnet launch, while the first half-hour exposed the practical gap between starting a chain and securing timely finality.