Foundry's Share of Bitcoin Hashrate Crosses 32%, Reviving Pool Centralization Debate

Foundry Digital's share of Bitcoin network hashrate crossed 32% for the first time in March, according to pool distribution data from mempool.space. The U.S.-based mining pool is now within distance of the informal 33% threshold that the Bitcoin protocol community has historically treated as the point at which pool centralization becomes a systemic concern. The milestone is reviving discussions — among protocol developers, pool operators, and miner customers — about how to reduce pool-level concentration without disrupting the operational plumbing that miners rely on.

Why 33% matters

The Bitcoin protocol is theoretically secure against any adversary controlling less than 50% of network hashrate. The 33% threshold is not protocol-level; it is a behavioral threshold. A mining pool controlling 33% can engage in selfish mining strategies — specifically, withholding blocks and releasing them strategically — that would be unprofitable at lower levels of control and that would be devastating at higher ones. Foundry, to be clear, is not engaging in anything of the sort. The concern is structural, not behavioral.

"The issue is not what Foundry is doing with 32%. The issue is that the industry has allowed any single pool to reach a position where doing something bad with it would be possible." — Matt Corallo, protocol engineer

Stratum V2 as structural response

Stratum V2 is the long-discussed update to the mining pool protocol that would allow individual miners to construct their own block templates rather than accepting templates provided by the pool operator. In the V1 system, the pool operator decides which transactions go into blocks. In V2, individual miners can make that decision, with the pool operator primarily coordinating payouts.

The difference matters because it changes the distribution of one specific power — the power to censor transactions — from the pool operator to the individual miners. Pool concentration becomes less meaningful under V2 because pools no longer control block content.

Adoption of V2 has been uneven. Several smaller pools, including Demand Pool and Ocean, have been running V2-native operations for over a year. The largest pools — Foundry, Antpool, ViaBTC — have not yet made the transition. The technical barriers are modest. The incentives are where the real question sits.

What the miner customers are saying

Marathon, Riot, and CleanSpark have all, in recent quarterly calls, been asked about their pool arrangements. The answers have been noncommittal. Several of the mid-sized miners, who do not have public earnings calls, have been more direct in private conversations: they prefer the operational simplicity of Foundry's current arrangement, and they are not motivated to switch unless either a cost advantage or a regulatory incentive appears.

Foundry itself has been publicly supportive of V2 adoption while moving deliberately. A switch to V2 for Foundry's entire customer base is a nontrivial engineering project, and the pool has not committed to a timeline.

The regulatory dimension

The 33% threshold is also a concern for the U.S. regulatory community, for reasons that are not primarily about Bitcoin. A single U.S.-domiciled entity processing 33% of blocks creates an unwanted centralization point for sanctions compliance, transaction censorship requests, and other jurisdictional pressures. Foundry has been reasonably transparent about how it handles such requests, but the existence of the possibility is itself the concern.

The practical path forward

A combination of V2 adoption, voluntary customer redistribution by larger miners, and pressure from developer communities will probably keep Foundry below 40% through the rest of the year. Whether the pool ends 2026 materially lower than 32% depends on whether the larger miners make the switch that they have, so far, declined to make. The incentives are slowly realigning. The question is whether they realign before an incident forces the issue.