Executive summary: context, proximate causes and implications

A brief chain split at height 941,881 produced a two‑block reorganization after competing miners produced closely timed blocks; the network ultimately adopted the longer sequence that came from a single dominant pool. The fork resolved within minutes and transactions from the discarded branch returned to mempools for reprocessing, but the event exposed the mechanics that allow a concentrated set of miners to determine which blocks persist. This episode arrived amid a pronounced downward difficulty adjustment and a multi‑month retreat in network processing power, compressing the pool of active operators and increasing the probability that a single actor strings together contiguous blocks.

On the ledger, the contest began when two large pools discovered valid headers within seconds; subsequent successive wins by one pool produced a heavier chain that eclipsed rival work and orphaned the competing blocks. Those orphaned blocks were not destroyed; their transactions were requeued, yet the miners who invested energy received no reward for that specific work. The scenario is a direct microcosm of structural concentration: as smaller or marginal miners exit because economics turn negative, remaining hashpower aggregates into fewer pools capable — by probability alone — of producing back‑to‑back wins and briefly reshaping canonical history.

Independent reporting and network telemetry point to a proximate operational shock that amplified the effect: regionally disruptive weather and grid stress temporarily pulled substantial capacity offline, and at least one major pool (Foundry USA) reported a sudden, large decline in contributed hashpower that materially shifted short‑term block production. Across the market, seven‑day averages show total network hashing falling below the 1,000 EH/s threshold and recent instantaneous readings clustered near the ~920 EH/s level cited here; other sources measuring slightly different windows report troughs nearer 913 EH/s. Block intervals lengthened during the outage window and briefly exceeded the design target, at times surpassing 11 minutes between blocks.

There are also competing accounts of the magnitude of the difficulty adjustment: on‑chain readings tied to our reported retarget show an approximate 7.76% decline, while other data providers and contemporaneous summaries reference larger single‑adjustment moves (figures reported in the low‑double digits) or project an additional roughly 10% change at the next retarget if outages persist. These differences reflect measurement timing, provider sampling windows and occasional unit reporting inconsistencies (terahashes vs. exahashes), not a cryptographic split — the network’s retarget mechanism behaved as designed, but near‑term statistics vary by source.

Market signals accompanying the technical event underscore economic stress: modelled production costs remain above spot prices for many operators, hashprice per unit of work has ticked modestly higher even as aggregate capacity contracts, and some operators are selling BTC reserves to cover expenses. A growing number of operators are evaluating or pursuing conversions of grid‑connected mining campuses into AI/HPC colocation to monetize stranded power and cooling — a structural reallocation that could permanently reduce spare proof‑of‑work capacity if executed at scale.

Operationally, the chain handled the disruption without protocol failure, but the incident serves as a real‑world demonstration that economic concentration has on‑ledger effects: short reorgs become more likely when diversity of block producers erodes. For custodians, exchanges and high‑volume services the response will likely be practical — raising confirmation counts or imposing asymmetric delays for on‑chain receipts — which in turn degrades the user experience and nudges settlement toward off‑chain or custodial pathways. In sum, the event is a near‑term warning shot: protocol rules held, but market structure and real‑world outages combined to produce measurable centralization risk and operational friction.