Insights › Mining Pools › Pool Review › 29 May – 18 Oct 2026
19 Oct 2026Mining Pools Review · Initiation of coverage
Bitcoin mining pools – last 10 epochs: F2Pool climbs as SpiderPool halves in the same week, Foundry USA trails the network
- Over 10 complete difficulty epochs (blocks 951,552 onwards, 29 May – 18 Oct 2026), capturing a couple of days of the current one, network hashrate rose +10.3%, from 926 to 1,022 EH/s (BMM node, 7-day estimate).
- F2Pool was the strongest and most consistent gainer: estimated hashrate +58.3% (107 → 169 EH/s), block share 11.5% → 16.5%. SpiderPool had the largest fall, -41.3% (101 → 59 EH/s), its share down from 10.9% to 5.8%. Most of the fall came in one week, 8–15 July, when SpiderPool lost 52 EH/s and F2Pool added 34 EH/s. BMM will keep monitoring whether this was a statistical coincidence or a lasting alignment of pools and/or miners.
- Foundry USA remained the leader but grew only +0.8%, 2 EH/s, against the network's +10.3%; its share of the network fell from 25.8% to 23.6%. Over the same period AntPool's share rose from 20.0% to 22.4% (added 44 EH/s) and ViaBTC's from 7.8% to 10.1% (added 30 EH/s).
- Concentration increased: the 3 largest pools found 63.7% of blocks in the last full epoch (from block 969,696 – 3 Oct), up from 56.6% in the first (from block 951,552 – 29 May). The top 5 pools (with ViaBTC and SpiderPool) rose from 76.5% to 80.2% of the network. Both figures are sums of the pool shares in the epoch table below. BMM's new BMMAA Pool Concentration Risk for the whole period is +0.68 (red alert): the top 5 pools added 82 EH/s while remaining pools fell by 22 EH/s.
- The first half of July was exceptionally eventful, including the seasonal peak of Texas grid curtailment, which affected a number of operations in the largest single region of hashrate concentration. The network's 7-day hashrate fell from 964 EH/s on 1 July to 900 EH/s on 15 July (-6.6%). The top 5 and all other pools fell -6.6% and -6.7% respectively, so BMMAA stayed neutral (+0.01).
BMM initiation of coverage
Pools remain one of the main centralisation points of the Bitcoin mining ecosystem, alongside hardware production. Currently pools decide how blocks are formed, including which transactions are selected. The BMM Mining Pools Review aims to build awareness of developments in the mining-pool ecosystem and of the technical changes that are likely to accelerate as the industry approaches the next halving cycle.
This is an initiation of coverage, not a full market review. It sets a baseline for the 5 largest pools by blocks found on BMM's own node, shows how hashrate flowed between them and the rest of the network over the last 10 difficulty epochs, and lists what BMM will watch next.
Top 5 pools: 10 full epochs + 2 full days of current (29 May – 18 Oct)
To read the graph: estimated hashrate, EH/s, 7-day basis. Upper band: network hashrate (axis cut between pools and network; right scale = change since start). Shaded columns alternate by difficulty epoch; the amber column marks 1–15 July. Click a pool name to hide or show it; hover for values and share of blocks.
Each pool line is the pool's share of blocks over the previous seven days multiplied by the network's 7-day hashrate. Block discovery is random, so single-day moves should not be over-read; the table below summarises the whole period.
| Pool | Start, EH/s | End, EH/s | Change | Average | High | Low | Share start → end | Blocks found |
|---|---|---|---|---|---|---|---|---|
| Foundry USA | 239.3 | 241.1 | +0.8% | 236.0 | 273.7 10 Oct | 198.5 13 Aug | 25.8% → 23.6% | 5,226 |
| AntPool | 185.0 | 228.7 | +23.6% | 178.6 | 229.7 16 Oct | 142.1 26 Aug | 20.0% → 22.4% | 3,966 |
| F2Pool | 106.6 | 168.7 | +58.3% | 139.5 | 184.2 11 Oct | 94.8 8 Jun | 11.5% → 16.5% | 3,095 |
| ViaBTC | 72.4 | 102.9 | +42.1% | 89.3 | 111.8 15 Oct | 72.4 4 Jun | 7.8% → 10.1% | 1,972 |
| SpiderPool | 100.6 | 59.1 | -41.3% | 81.9 | 117.0 18 Jun | 46.3 16 Oct | 10.9% → 5.8% | 1,813 |
| All other pools | 222.2 | 221.1 | -0.5% | 203.7 | 250.8 27 Jun | 164.1 23 Jul | 24.0% → 21.6% | 4,518 |
| Network (BMM node) | 926.2 | 1,021.5 | +10.3% | 928.9 | 1,030.2 15 Oct | 853.1 9 Jun | 100% | 20,590 |
Estimated hashrate, 7-day basis: start 4 Jun, end 18 Oct 2026. High and low over the period. Blocks found: 29 May – 18 Oct 2026 (full UTC days).
Closer look: 1–15 July
The first half of July was exceptionally eventful. The seasonal peak of Texas grid curtailment affected a number of operations in the largest single region of hashrate concentration: ERCOT demand-response and summer peak-load programmes lead large miners to curtail on the hottest days, which reduces running load temporarily rather than removing machines for good. The network's 7-day hashrate fell from 964 EH/s on 1 July to 900 EH/s on 15 July (-6.6%), and the difficulty retarget on 11 July fell 5.00%, the second-largest drop in the period.
Price was not the trigger. Bitcoin closed June at $58,532 and 15 July at $64,709, and hashprice (with fees) rose from about $28.4 to $32.3 per PH/s per day over the same fortnight (BMM data). On top of the curtailment, SBI Crypto announced on 2 July that its pool would close by the end of the month, giving its miners a reason to choose a new pool during the same weeks and resulting in roughly 10 EH/s moving to other pools during July (SBI's seven-day average fell from 16.2 EH/s on 30 June to 5.8 EH/s on 30 July, CryptoSlate).
Within the top 5, the moves were very uneven, with SpiderPool losing 56 EH/s and AntPool losing 22 EH/s, while F2Pool added 18 EH/s, ViaBTC 7 EH/s and Foundry USA 4 EH/s during those 15 days. As a group the top five lost 49 EH/s (-6.6%) and all other pools 15 EH/s (-6.7%). Measured with the BMMAA Pool Concentration Risk, which compares the two growth rates on a scale from −1 to +1, the fortnight scores +0.01: neutral. The near-zero reading is the point: the drop was not smaller pools losing ground but the largest miners, connected to the largest pools, curtailing in step with the network.
Dynamic: Pools vs Network
With the network up +10.3% over the period, 3 growth leaders were F2Pool (+58.3%), ViaBTC (+42.1%) and AntPool (+23.6%) that outgrew the network. Foundry USA (+0.8%) and the group of all remaining pools (-0.5%) did not, and SpiderPool fell sharply. In absolute terms F2Pool added 62 EH/s and AntPool 44 EH/s, together more than the whole network's net growth, while SpiderPool gave up 41 EH/s.
Where one pool's gain lines up in time with another's loss, as F2Pool's and SpiderPool's did in mid-July, the pattern is consistent with miners redirecting hashrate between pools. It remains an indication until the pools or the miners confirm it: block attribution shows where blocks were found, not which miners connected where. The same caution applies to the two closures in the period; neither SBI Crypto nor Binance Pool is large enough to explain the top-five moves on its own.
Share of blocks by epoch
| Epoch from block | Start | Retarget | Foundry USA | AntPool | F2Pool | ViaBTC | Spider Pool | Other | Top 5 | Top 5, EH/s | Network, EH/s | Top 5 Δ / other Δ, EH/s | BMMAA |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 971,712 * | 16 Oct | +5.92% | 22.9% | 21.0% | 17.2% | 9.5% | 7.0% | 22.4% | 77.6% | 796 | 1,026 | +3 / +34 | -0.61 Deep blue |
| 969,696 | 3 Oct | -0.03% | 26.2% | 20.6% | 16.9% | 10.4% | 6.0% | 19.8% | 80.2% | 793 | 990 | +44 / +2 | +0.28 Neutral |
| 967,680 | 19 Sep | +4.16% | 24.6% | 21.4% | 15.9% | 9.7% | 7.8% | 20.6% | 79.5% | 750 | 944 | +16 / -16 | +0.49 Amber |
| 965,664 | 5 Sep | +1.31% | 25.2% | 19.1% | 14.8% | 10.2% | 8.5% | 22.2% | 77.8% | 734 | 943 | +47 / -6 | +0.49 Amber |
| 963,648 | 23 Aug | -1.31% | 25.8% | 17.4% | 14.8% | 8.6% | 9.5% | 23.9% | 76.1% | 686 | 902 | -27 / +19 | -0.57 Blue |
| 961,632 | 8 Aug | +0.99% | 24.2% | 20.4% | 15.3% | 9.2% | 9.3% | 21.6% | 78.4% | 713 | 910 | +10 / -3 | +0.23 Neutral |
| 959,616 | 25 Jul | -0.74% | 25.8% | 17.2% | 17.3% | 9.1% | 8.5% | 22.1% | 77.9% | 703 | 903 | -8 / +14 | -0.46 Blue |
| 957,600 | 11 Jul | -5.00% | 27.4% | 18.6% | 16.8% | 9.7% | 6.8% | 20.7% | 79.3% | 711 | 897 | -38 / -16 | +0.12 Neutral |
| 955,584 | 27 Jun | +7.15% | 25.1% | 19.0% | 13.6% | 10.4% | 10.7% | 21.2% | 78.8% | 749 | 951 | +37 / -20 | +0.58 Amber |
| 953,568 | 14 Jun | -10.09% | 25.0% | 19.0% | 12.3% | 9.0% | 10.9% | 23.7% | 76.3% | 712 | 934 | +2 / +3 | -0.11 Neutral |
| 951,552 | 29 May | +1.72% | 25.2% | 19.4% | 12.0% | 9.5% | 10.5% | 23.5% | 76.5% | 711 | 929 | – | – |
Share of blocks found in each epoch (UTC days from the retarget date). * current epoch, still open. Retarget: difficulty change at the start of the epoch (red = up, blue = down). Top 5: the 5 pools in this review together. Latest epoch first. BMMAA: BMMAA Pool Concentration Risk for the epoch against the previous one, from −1 to +1; within ±0.30 neutral (block luck), above +0.30 concentrating, below −0.30 decentralising. Method: BMMAA.
BMMAA in the epoch table turns this into one number per epoch. Above +0.30 the top five grew clearly faster than smaller pools and concentration rose; below −0.30 smaller pools gained; in between, the moves are within the range block luck alone produces. Concentration rose clearly in 3 of 9 steps between full epochs and fell clearly in 2. The clearest absorption came in the epoch from 27 June: the network added 17 EH/s, but the top five added 37 EH/s while other pools lost 20 EH/s (BMMAA +0.58). In the epoch from 19 September the network was flat, yet the top five gained 16 EH/s at the expense of smaller pools (+0.49). The epoch from 23 August went the other way: the network fell 8 EH/s, yet smaller pools grew 19 EH/s while the top five lost 27 EH/s (-0.57), a clear decentralisation move. BMM now publishes BMMAA daily on 30-day windows (latest +0.61, Red alert · strong concentration); see BMMAA Pool Concentration Risk.
Leaderboard: Top 3 – concentration risk above 51%
Estimated hashrate of the three largest pools, EH/s (left), and network hashrate change since the start of the period, % (right, dashed). Hover for values.
Foundry USA, AntPool and F2Pool together found 63.7% of blocks in the last full epoch, well above the 51% threshold for any single pool or coordinated group of pools. No pool is close to that level on its own, but the trend is towards fewer, larger template builders, which is why protocol work that lets miners build their own templates matters (see Stratum V2 below).
Foundry USA has led throughout, but its margin has narrowed. Since Binance's closure announcement on 13 October its 7-day share stands at 23.6% against AntPool's 22.4% and F2Pool's 16.5%. F2Pool's rise was the steadiest of the three: its share was higher in the last full epoch than in the first, with no large one-off step apart from July. AntPool's path was more volatile, with a dip in July and a strong recovery from August.
Foundry USA serves mainly North American miners. That is where listed mining companies are converting power and sites to AI and high-performance computing: the tracked public-miner cohort's reported hashrate fell 13.4% from Q4 2025 to Q2 2026, faster than the network over the same quarters (TheEnergyMag). Foundry trailing the network over the period is consistent with that shift, but pool data cannot show where individual miners operate, so BMM treats this as a working hypothesis to test in future editions.
SpiderPool, the fifth pool in this review, was in the top 5 for the whole period and is included in all figures. In the data from 12–18 October its share fell sharply and it is no longer number 5 on recent days; BMM will update the set in the next review. With Binance Pool winding down, BMM expects Luxor to climb into the top five unless there is a major reshuffle.
Key developments in the mining-pool market
SBI Crypto closes its pool
SBI Crypto announced on 2 July 2026 that its Bitcoin pool would stop accepting shares and close, asking miners to point their machines to another pool before the cut-off. Mining connections were disconnected from 22:00 UTC on 30 July, and the Stratum server was switched off at 23:59:59 UTC that day. At the time of the announcement the pool was reported at roughly 2% of network hashrate.
CryptoSlate's tracking shows the pool emptying gradually rather than on the final day: SBI's seven-day average fell from 16.2 EH/s on 30 June to 5.8 EH/s on 30 July, so most of its miners had moved during July. In BMM's data no single top-five pool picked up a matching share in the two weeks after the cut-off: all other pools went from 21.7% to 22.2% of blocks, while Foundry USA's share fell from 27.3% to 23.6%.
CoinDesk, 2 July 2026 ↗ · CryptoSlate ↗ · SBI Crypto ↗
Binance Pool to close on 10 November
Binance announced on 13 October 2026 that Binance Pool and related services will be discontinued. At 07:00 UTC on 10 November the pool stops accepting shares and its Stratum endpoints are disabled; Pool Savings, cloud mining, the transaction accelerator and the pool APIs close with it. Binance asked miners to reconfigure their machines to another pool before that date and said the pool would run normally until the cut-off.
Binance Pool sits in the "all other pools" group in this review. Its closure is a clearly dated test of where pool hashrate goes when a provider exits. BMM has recorded the pre-closure baseline in this edition and will compare the top five before and after 10 November in the next one.
Stratum V2: first Job Declaration block and a growing working group
DMND reported that block 955,318, mined on 26 June 2026, was the first Bitcoin block produced with Stratum V2 Job Declaration. With Job Declaration the miner, not the pool, builds the block template and selects the transactions; the pool still coordinates the work and the payouts. The block gives the protocol a specific on-chain reference point; it does not mean that all DMND miners, or the wider network, use the feature.
The Stratum V2 Working Group welcomed AntPool, Block, F2Pool, Foundry, MARA Foundation, SpiderPool and DMND as new members on 7 May 2026, so 4 of the 5 pools in this review now take part. Auradine is listed as supporting Stratum V2 across its ASICs through FluxOS; however, Auradine has since pivoted to AI under the new name Velaura. Wider use could make block attribution more diverse, as more miners build their own templates and some choose not to tag their blocks. For readers who want an introduction, F2Pool's Stratum V2 podcast with Pavlenex is a good place to start.
Stratum V2 ↗ · New working-group members ↗ · DMND ↗ · Braiins ↗ · Spiral ↗ · AntPool ↗ · Block ↗ · F2Pool ↗ · Foundry ↗ · MARA Foundation ↗ · SpiderPool ↗ · F2Pool podcast on Stratum V2 ↗ · Pavlenex on X ↗
CoinEx wind-down and ViaBTC
CoinEx announced on 15 September 2026 that it would cease operations through an orderly wind-down: Earn, staking and other non-spot products stopped on 22 September, spot trading on 29 September, and withdrawals remain available until 22 December 2026. On the same day ViaBTC announced that its automatic withdrawal to CoinEx would end on 22 September, and later stated that its pool is independent of CoinEx, with a separate legal entity, accounts, infrastructure and servers.
For pool users the practical point was the loss of a connected payout destination and the need to update withdrawal addresses; the notices do not indicate that ViaBTC's mining service was affected. ViaBTC's estimated hashrate was in fact +42.1% higher at the end of the period than at the start.
ViaBTC notice ↗ · ViaBTC clarification ↗ · CoinEx announcements ↗
OCEAN: leadership change after BIP-110
OCEAN and Luke Dashjr announced on 29 August 2026 that they had agreed to separate, with Dashjr leaving his roles as chairman, chief technology officer and director. The joint statement cited differing visions for the future of Bitcoin mining after recent protocol developments. The separation followed the controversy around BIP-110, a proposal to temporarily restrict some forms of non-financial data in Bitcoin transactions, whose activation attempt divided opinion on protocol policy and on the role of pools in block construction.
How the change came about remains a matter of debate, and BMM does not take a view on it; time will show. Under its new leadership OCEAN has raised its public profile, and BMM will track whether that turns into a visible change in OCEAN's block share.
Joint statement ↗ · BIP-110 explainer ↗
Neopool security incident
Neopool said it detected an information-security attack on 29 July 2026. In a statement on 7 August it reported that attackers had taken control of the neopool.com domain, its e-mail service and parts of its infrastructure, that incident response had started and that law enforcement was involved. It said user funds remained secure, warned users to treat @neopool.com e-mails as untrusted, and recommended moving hashrate to another pool while it recovered.
A later update on its Telegram channel reported that the domain and e-mail had been restored, user payments made in full, and that normal operations would resume after infrastructure upgrades. The episode is a reminder that a pool's web and e-mail infrastructure is part of the security perimeter for its miners.
Neopool statement ↗ · CryptoNews, 7 August 2026 ↗
Poolin and the US bankruptcy filing
Poolin-related US entities filed for Chapter 11 in July 2026 over their remaining Texas mining assets. The filings concern those entities and sites, not by themselves the Poolin pool service, although they prompted speculation about the pool.
BMM's estimate is that much of the hashrate attributed to Poolin comes from its own mining operations; this remains a BMM estimate until fleet disclosures confirm it. BMM will watch whether Poolin's attributed block production changes as the sale proceeds.
The Block ↗ · Data Center Dynamics ↗
Oman's state-backed pool initiative
In June 2026 Enegix Global announced that it was providing the technology platform and liquidity infrastructure for Omanhash.om, described as Oman's national Bitcoin mining-pool initiative and associated with the country's Ministry of Transport, Communications and Information Technology. It was presented as infrastructure for licensed mining operators in Oman, and as Enegix's second sovereign mandate after Kazakhstan.
The Omanhash.om website is no longer available, and BMM has not seen any later operational milestones or identifiable block production. That does not prove the project has ended; BMM will note any progress in future editions.
Enegix Global announcement ↗ · Bitcoin.com News ↗
Solo mining and small devices
Bitaxe and similar open-source home miners add very little to network hashrate, but they matter beyond their size. They bring people into hands-on mining, support experiments with open firmware and hardware, and keep the community involved in discussions about pools, block templates and the protocol. Solo and community pools give these miners a home, and each block they find gets wide attention.
A block found by a small miner is a rare event and says nothing about network hashrate, and the blockchain does not show which device found a block. BMM sees the growth of this segment as a healthy sign of participation.
Solo CKPool ↗ · Bitaxe Pool ↗ · Public Pool ↗
North American miners and the AI pivot
TheEnergyMag's comparison of listed miners reports realised hashrate of 368.3 EH/s in Q4 2025, 344.4 EH/s in Q1 2026 and 319.0 EH/s in Q2 2026, a 13.4% fall, while the network's quarterly average fell 10.6% over the same quarters. A separate TheEnergyMag article put the first-half reduction at about 75 EH/s, as power and sites were converted to AI and high-performance computing.
These are figures for publicly reporting miners, not a measure of hashrate permanently lost: machines may be idled, sold or moved elsewhere. For pools, the question is whether those whose clients are mostly listed North American miners lose share as fleets age and are not replaced. That is the hypothesis behind the Foundry USA observation above, and BMM will test it edition by edition.
TheEnergyMag, 1 October 2026 ↗
Russia: connectivity and pool choice
Some pools withdrew from Russia after 2022 for fear of sanctions, and Russia has since expanded traffic filtering at internet-provider level. Some Russian operations that pointed hashrate to pools abroad now find those destinations affected by ISP filtering. BMM will record documented changes with dates and sources; pool data alone cannot show a miner's location, so any link to pool shares will be made only where the evidence supports it.
What BMM will watch next
With the halving cycle now past its midpoint, the pressure on miners will build. Large new capital deployment is likely to slow sharply, because the time left to recoup an investment before the next halving keeps shrinking. In that setting BMM expects buyers of Bitmain equipment to be encouraged to point their hashrate to AntPool, so AntPool's hashrate should stay firm and is likely to grow.
Foundry USA's leadership is likely to be challenged soon. Its concentration in North America, where listed miners are pivoting to AI, is becoming a pressure point. BMM takes a positive view of F2Pool, which has grown consistently over long periods, step by step, and is likely to keep growing. Together with ViaBTC it may capture most of the hashrate redistributed from Binance Pool.
The second tier, the pools just below the top 5, will face more competition as they fight for growth and try to move closer to the top. At the same time BMM hopes for a shift towards decentralisation, which is why it introduced the BMMAA Pool Concentration Risk to track the direction every day. BMM will watch how Stratum V2 deployment takes off at pools in the coming months.
Another major reset BMM expects is in the physical location of hashrate. It will change the paradigm for the many pools that kept their main focus on the US while Texas and other locations were building out. On its side, BMM will keep adding granularity to its data and analytics, and will roll out further updates to the dashboard and to BMM Research & Analysis.
Sources
- BMM Bitcoin network dashboard ↗
- CoinDesk: SBI Crypto to shut down mining pool (2 July 2026) ↗
- CryptoSlate: SBI Crypto pool closure and pool concentration ↗
- Binance: Binance Pool discontinuation (13 October 2026) ↗
- Stratum V2 Working Group: new members (7 May 2026) ↗
- ViaBTC: discontinuation of withdrawal to CoinEx ↗
- Joint statement of OCEAN and Luke Dashjr (29 August 2026) ↗
- Neopool: information-security incident ↗
- The Block: Poolin files Chapter 11 ↗
- Enegix Global: Omanhash.om announcement ↗
- TheEnergyMag: Bitcoin's Great Unplug (1 October 2026) ↗
Data and method
Blocks by pool come from BMM's own Bitcoin node (coinbase attribution), 29 May – 18 Oct 2026, full UTC days (20,590 blocks). A pool's estimated hashrate is its share of blocks over the previous seven days multiplied by the network's 7-day hashrate estimate from the same node; the line therefore starts seven days into the period. Epoch shares count blocks found on UTC days from each retarget date. BMMAA per epoch is the change in the top-five share of blocks against the previous epoch (pp), and in EH/s the top five's net change minus their fair share of the network's change; the daily 30-day version and its limits are described on the BMMAA methodology page. Block discovery is random, so short-window shares carry noise. Event dates and pool news are from the sources listed and were not independently verified by BMM.
For information only; not investment advice. Figures are estimates and may be revised as data is updated.
