Bitmain's Grip on Bitcoin: ASIC Manufacturing Concentration and Its Network Risks

Bitcoin's security depends on miners competing to produce valid blocks. But the hardware those miners run comes from a remarkably small number of manufacturers. Bitmain, a single Chinese company, controls an estimated 50 to 65 percent of the global ASIC miner market. Add MicroBT and Canaan, and three companies account for over 90 percent of all Bitcoin mining hardware sold worldwide. This level of concentration in mining hardware manufacturing creates supply chain risks that most discussions of Bitcoin decentralization overlook.

Decentralization at the protocol layer means little if a handful of companies control the physical machines that produce hashrate. Firmware backdoors, chip fabrication dependencies, geopolitical exposure, and tariff shocks all trace back to the same structural problem: the ASIC manufacturing market is an oligopoly, and Bitmain sits at its center.

Who Makes Bitcoin Miners

The Bitcoin ASIC market was valued at approximately $11.4 billion in 2025 and is projected to reach $12.4 billion in 2026. Three manufacturers dominate this market, each with distinct supply chain relationships and geographic footprints.

Bitmain

Founded in 2013 by Jihan Wu and Micree Zhan, Bitmain manufactures the Antminer series and designs its own BM-series ASIC chips. The company holds an estimated 50 to 65 percent of global market share, though exact figures are difficult to verify because Bitmain remains private after its Hong Kong IPO application lapsed in March 2019. A protracted power struggle between its co-founders from 2019 to 2021, which included physical confrontations at company offices, ended with Zhan regaining control and Wu departing. Bitmain fabricates its leading-edge chips at TSMC on the 5nm process node.

MicroBT

Founded in 2016, MicroBT produces the Whatsminer series and holds roughly 20 to 30 percent of the market. The company is particularly strong in North American deployments. Unlike Bitmain, MicroBT uses Samsung Foundry for chip fabrication: its M56S++ was the first high-volume Bitcoin ASIC built on Samsung's SF3E 3nm gate-all-around process.

Canaan

Canaan shipped the first commercial Bitcoin ASIC (the Avalon 1) in 2013 and trades publicly on NASDAQ. Despite its pioneering role, Canaan holds only 5 to 10 percent of the market. The company has narrowed the efficiency gap with its Avalon A16 series, launched at Blockchain Life 2025, which delivers 300 TH/s at 12.83 J/TH: competitive with Bitmain's best air-cooled machines.

Market share caveat: None of the three major ASIC manufacturers publish audited market share data. Estimates vary depending on whether the measurement is units shipped, revenue, or deployed hashrate. The ranges cited here are compiled from industry analyst reports and should be treated as approximations.

Current-Gen ASIC Efficiency Comparison

The efficiency race in Bitcoin mining hardware has pushed power consumption below 13 J/TH for the most advanced models. This table compares flagship ASICs from all three manufacturers across air-cooled and immersion/hydro cooling configurations. Efficiency (measured in joules per terahash) is the single most important metric for mining profitability: lower is better.

Bitmain consistently leads on efficiency in both air-cooled and hydro-cooled categories, though Canaan's A16 XP has closed the gap significantly. MicroBT's strength lies in immersion cooling configurations, where its Samsung 3nm chips deliver competitive performance at scale. All three manufacturers are racing toward 2nm process nodes for their next generation of chips.

The Chip Fabrication Bottleneck

Every Bitcoin ASIC depends on advanced semiconductor fabrication that only two foundries in the world can provide at scale. This creates a critical chokepoint in the mining hardware supply chain.

The foundry split is notable: Bitmain relies exclusively on TSMC, while MicroBT and Canaan use Samsung. China's SMIC has been targeting Bitcoin mining ASICs on its 7nm process, positioning itself as a domestic alternative, but it remains multiple generations behind TSMC and Samsung in efficiency. Each process node improvement delivers roughly 25 to 40 percent better energy efficiency per hash, making foundry access a decisive competitive advantage.

This dual-foundry dependency exposes Bitcoin's hashrate to risks that have nothing to do with the protocol itself. A disruption at TSMC (whether from natural disaster, geopolitical conflict around Taiwan, or capacity allocation decisions favoring higher-margin AI chips) could constrain Bitmain's supply. Similarly, Samsung's foundry decisions affect both MicroBT and Canaan simultaneously.

The Graveyard of ASIC Competitors

The concentration we see today is partly a result of how difficult it is to compete in ASIC manufacturing. Several companies have tried and failed, reinforcing the incumbents' market position.

Halong Mining

Halong emerged in 2018 with the Dragonmint T1, which used Samsung chips and was marketed as a credible Bitmain alternative. The company's last social media activity was May 2018, after which it ceased production entirely. Questions about whether the Dragonmint T1 was a rebranded Innosilicon device were never fully resolved.

Ebang

Ebang released the Ebit E11++ (44 TH/s) in 2018 and successfully listed on NASDAQ in June 2020 after a failed Hong Kong IPO. By June 2022, the company received a NASDAQ delisting warning when its share price fell below $1.00. While Ebang avoided delisting, it has not released competitive mining hardware in years and has effectively exited the ASIC manufacturing market.

Innosilicon

Known for its Terminator series (SHA-256) and A10/A11 Pro (Ethash), Innosilicon was hit hard by Ethereum's merge to proof-of-stake in September 2022, which rendered its Ethash miners obsolete overnight. Firmware bugs, inconsistent build quality, and an inability to secure comparable foundry relationships with TSMC or Samsung eroded its competitive position. The company has largely exited the market without a formal shutdown.

Pattern to notice: Every failed competitor struggled with the same barriers: securing advanced chip fabrication capacity, building reliable firmware, and establishing distribution networks. These barriers to entry naturally concentrate the market around well-capitalized incumbents, making the oligopoly self-reinforcing.

Firmware Risks: Antbleed and Covert ASICBoost

When a single manufacturer controls the majority of deployed mining hardware, firmware becomes a systemic risk vector. Bitmain's history provides two documented examples.

The Antbleed Backdoor

Discovered in April 2017, Antbleed was a backdoor found in the stock firmware of all Antminer machines. The firmware contacted a Bitmain central service at random intervals between 1 and 11 minutes, transmitting each miner's serial number, MAC address, and IP address. The mechanism gave Bitmain the theoretical ability to remotely shut down any Antminer or redirect its hashrate.

At the time of discovery, Bitmain hardware was estimated to produce around 70 percent of the Bitcoin network's hashrate, meaning this backdoor could theoretically have disabled the majority of the network's mining capacity. Bitmain claimed the feature was unfinished and was intended as a tool for customers to recover stolen miners. Bitcoin Core developer Luke Dashjr characterized it as "not a bug, but intentional malware."

Covert ASICBoost

ASICBoost, a technique documented by researchers Timo Hanke and Sergio Lerner, exploits a mathematical shortcut in SHA-256 computation to achieve a 10 to 20 percent increase in mining efficiency. Bitmain deployed a "covert" version of ASICBoost by grinding merkle roots instead of nonces, gaining an undisclosed competitive advantage over other miners.

The controversy extended beyond efficiency gains. Covert ASICBoost was linked to Bitmain's opposition to SegWit, because SegWit's transaction format changes would have broken the covert implementation. This meant a hardware manufacturer's proprietary optimization was influencing Bitcoin's protocol governance: a concrete example of how manufacturing concentration can affect consensus-level decisions.

Geopolitical Exposure: Tariffs and Trade Wars

All three major ASIC manufacturers are headquartered in China, creating geopolitical risk that became acute in 2025 when US tariffs on Chinese imports reached 145 percent. Tariffs on ASIC miners specifically jumped from 2.6 percent to over 21 percent, with some categories facing rates above 150 percent.

The industry's response was immediate. US miners imported $860 million worth of mining equipment in Q1 2025 alone, racing to beat tariff deadlines. Industry sources compared the potential disruption to China's 2021 mining ban, with breakeven costs pushed above $80,000 per BTC for some operators.

Manufacturing Relocation

All three manufacturers have responded by moving production outside China:

• Bitmain opened its first US assembly line in January 2025 and also manufactures in Malaysia, Thailand, and Indonesia
• MicroBT shifted manufacturing capacity to Malaysia, Thailand, and Indonesia
• Canaan ships from production centers across North America, East Asia, and Southeast Asia

However, relocation addresses assembly, not chip fabrication. The silicon itself still comes from TSMC in Taiwan and Samsung in South Korea. US semiconductor export controls, while primarily aimed at AI and supercomputing applications, create an overhang of regulatory uncertainty for any company dependent on advanced process nodes.

The Open-Source Response: Bitaxe and Community Mining

The most promising counterweight to ASIC manufacturing concentration is the open-source hardware movement, led by the Bitaxe project. Bitaxe is the first fully open-source hardware Bitcoin ASIC miner: its designs, schematics, and firmware are all publicly available. The devices use genuine Bitmain ASIC chips (sourced from the secondary market) mounted on open-source PCBs with WiFi connectivity.

As of late 2025, open-source home mining hardware has mined at least five confirmed Bitcoin blocks, with three solved by Bitaxe devices and two by NerdQAxe++ miners. The first was block #853,742 on July 24, 2024. A stock Bitaxe Gamma running at 1.2 TH/s won block #889,975 in March 2025, earning its operator 3.149 BTC. The interval between confirmed solo blocks has been shrinking (229 days, 179 days, 52 days, 25 days), reflecting a growing installed base.

Bitaxe does not solve the concentration problem at its root. These devices still use Bitmain-designed chips, and their combined hashrate is negligible compared to industrial mining farms. But they serve a different purpose: proving that open-source mining hardware is viable, building a community of developers who understand ASIC design, and creating a foundation for future open-source chip development. Projects like NerdAxe, Ember One, and Piaxe extend this ecosystem with alternative form factors.

What Concentration Means for Bitcoin Security

The risks of ASIC manufacturing concentration are not hypothetical. They manifest across several dimensions that directly affect base layer security.

Supply Chain Disruption

If TSMC experiences a major disruption (earthquake, geopolitical conflict, capacity reallocation to AI chips), Bitmain cannot produce new ASICs. With 50 to 65 percent market share, this would constrain the supply of new mining hardware globally, potentially slowing hashrate growth or even causing hashrate decline as older machines reach end-of-life. The difficulty adjustment mechanism protects block production, but reduced hashrate lowers the cost of a 51% attack.

Firmware as Attack Surface

The Antbleed incident demonstrated that firmware-level backdoors in dominant mining hardware are technically feasible. While Antbleed was discovered before any known exploitation, the broader risk remains: stock firmware on the majority of active ASICs is closed-source and controlled by the manufacturer. Open-source firmware alternatives like Braiins OS mitigate this risk for operators who install them, but most deployed hardware runs manufacturer firmware.

Governance Influence

The covert ASICBoost episode showed that a dominant manufacturer's proprietary hardware optimizations can influence protocol-level debates. Bitmain's opposition to SegWit was not purely ideological: it was motivated by the fact that SegWit would neutralize a competitive advantage embedded in their hardware. When a single company produces most of the network's mining equipment, its commercial interests can distort soft fork activation dynamics.

Mining Pool Interaction

Manufacturing concentration compounds with mining pool centralization. Bitmain has historically operated major mining pools (Antpool, BTC.com), meaning the same entity that manufactures the hardware also directs hashrate. While Stratum V2 and decentralized pool protocols aim to separate block template construction from pool operators, progress has been gradual.

Emerging Countermeasures

Several developments are working to reduce manufacturing concentration risk, though none offer a complete solution in the near term.

• Canaan's A16 series has closed the efficiency gap with Bitmain, giving large-scale miners a more competitive second source
• MicroBT's Samsung relationship provides foundry diversification away from TSMC, reducing single-foundry risk
• Open-source firmware projects (Braiins OS, AxeOS) allow operators to replace manufacturer firmware, mitigating backdoor risk
• The Bitaxe community is building institutional knowledge in open-source ASIC design that could eventually enable independent chip development
• Geographic diversification of assembly (US, Malaysia, Thailand) reduces tariff exposure, even if chip fabrication remains concentrated

The fundamental tension: ASIC manufacturing is a capital-intensive business with extreme economies of scale. The same physics that make ASICs efficient at mining (custom silicon optimized for a single algorithm) also make them expensive to design and fabricate. This inherently favors large, well-funded incumbents. Any serious effort to decentralize mining hardware manufacturing must contend with this economic reality.

Why This Matters for Bitcoin Layer 2s

Every Bitcoin Layer 2 protocol depends on the base layer for final settlement. When users exit from Spark, Lightning channels, or any other off-chain system, they broadcast transactions to Bitcoin L1 and rely on miners to include those transactions in blocks. If mining hardware concentration leads to hashrate instability, supply chain disruptions, or governance capture, the settlement guarantees that L2 protocols depend on are weakened.

This is not a reason to avoid Layer 2 solutions. Rather, it underscores why the Bitcoin ecosystem needs to monitor and address concentration risks at every layer of the stack: from chip fabrication to mining pools to node software. For developers building on Bitcoin, understanding these supply chain dynamics is as important as understanding the protocol itself. The Spark documentation covers how Spark's exit mechanism ensures users can always settle to L1, but that exit path is only as robust as the base layer that processes it.

For a deeper look at Bitcoin's mining landscape, see our analyses of mining economics in 2026 and the security budget debate.

This article is for educational purposes only. It does not constitute financial or investment advice. Bitcoin and Layer 2 protocols involve technical and financial risk. Always do your own research and understand the tradeoffs before using any protocol.