Efficiency in Bitcoin mining has evolved significantly over the years, with technological advancements driving remarkable improvements. However, the pace of efficiency gains has slowed in recent years due to challenges in reducing semiconductor sizes and the physical limitations of current hardware. Cooling technologies and custom firmware have become essential tools for further optimization. Let’s delve into the evolution and future of efficiency in Bitcoin mining.
Efficiency in Bitcoin Mining
The Evolution of Efficiency in Bitcoin Mining
The Pace of Efficiency Gains is Slowing
Challenges of Reducing Semiconductor Sizes
Cooling Efficiency
Can We Expect Sub-10 J/TH Efficiency in 2025?
Premium Members Only:
Optimizing Efficiency with Custom Firmware
Normal Efficiency Mode (NEM) & High Efficiency Mode (HEM)
Importance of Management and Monitoring Software
Optimizing for Efficiency Doesn’t Always Lead to the Quickest ROI
Efficiency in Bitcoin Mining
Efficiency in Bitcoin mining refers to the ratio of power consumption to hashrate, indicating how much energy is required to perform a given amount of computational work. It is a key metric used to assess the performance and profitability of mining hardware.
Efficiency is calculated by dividing the power consumption (Watts) by the hashrate output of a machine or mining fleet (TH/s). The result is expressed in joules per terahash (J/TH). Lower efficiency values (J/TH) indicate a more energy-efficient machine, meaning it consumes less electricity per unit of hashrate produced.
The Evolution of Efficiency in Bitcoin Mining
The first Bitcoins were mined on January 3, 2009, using a CPU (Central Processing Unit). As the network grew, miners needed more computational power to solve a block before others could. It wasn’t until 2013 that the first ASIC miner, the air-cooled Avalon-1, was introduced. This machine produced 66 GH/s (0.066 TH/s) while consuming 600 watts, resulting in an efficiency of 9,090 J/TH.
In the second half of 2024, Bitmain released the air-cooled Antminer S21XP. This miner delivers 270 TH/s with a power consumption of 3,645 watts, achieving an efficiency of 13.5 J/TH. In just over a decade, hashrate output has increased by more than 4,000%, while efficiency has improved by 99%.

The Pace of Efficiency Gains is Slowing
Over the last few years, the pace of efficiency improvements in ASIC mining hardware has noticeably slowed. In the early years of Bitcoin mining, each new generation of hardware brought massive leaps in both hashrate and energy efficiency. However, as ASIC technology has matured, the rate of improvement has diminished. While early ASICs saw efficiency gains of 5x–10x per generation, more recent models offer modest improvements compared to previous generations.

Challenges of Reducing Semiconductor Sizes
The slowdown in efficiency gains is partially due to the fact that ASIC chips are nearing the physical limits of semiconductor technology, particularly as chip sizes have decreased significantly over time. Early ASIC miners were built with chips using a 55nm (nanometer) manufacturing process, which means the transistors on the chip were 55 nanometers apart. Over the years, advancements in semiconductor manufacturing have led to progressively smaller chips, with modern ASICs now using chips built with 3nm process technology.

The smaller the chip, the more transistors can be packed into the same area, allowing for increased computational power and efficiency. However, reducing chip sizes comes with significant challenges, both in terms of performance and reliability. Smaller chips are more prone to failure due to several factors. As transistors shrink, quantum effects like electron tunneling can cause reliability issues and performance degradation. Additionally, the increased number of transistors packed into a smaller space generates excess heat, which can lead to instability, thermal throttling, or even permanent damage. The smaller the chip, the less tolerance there is for manufacturing defects, which increases the likelihood of malfunctions and reduces yields. Finally, while smaller chips are designed to be more power-efficient, the closer transistor spacing complicates power management, making the chips more vulnerable to failures from minor power fluctuations.
Cooling Efficiency
Cooling plays a vital role in maintaining mining hardware efficiency by preventing overheating, which can lead to performance degradation and increased power consumption. Advanced cooling solutions such as immersion cooling and hydro cooling play an important role in pushing the efficiency boundaries. Immersion-cooling involves submerging ASICs in dielectric fluid, which efficiently absorbs and dissipates heat, reducing power consumption. Hydro-cooling, where water is circulated through specialized cooling plates, provides an effective means of heat dissipation.

Liquid-cooled technologies enable higher hashrate density and increased power input while maintaining the same efficiency levels as air-cooled mining equipment. They also minimize energy waste from excessive fan usage and extend the lifespan of ASIC miners, reducing long-term capital replacement costs. Despite these advantages, air-cooled mining remains popular because immersion and hydro-cooling setups require substantial infrastructure investment.
Can We Expect Sub-10 J/TH Efficiency in 2025?
As ASIC mining technology continues to evolve, the prospect of achieving sub-10 J/TH efficiency by 2025 is becoming increasingly realistic. Recent models are approaching 12–15 J/TH, further reductions in energy consumption are becoming harder to achieve due to the physical limitations of semiconductor technology. Nonetheless, advancements in cooling systems, chip design, and power management help push efficiency closer to this milestone.

The latest Bitcoin mining chip, SEAL03, has successfully completed testing following its tape-out. It achieved an impressive power efficiency ratio of 9.7J/TH while operating in low voltage, ultra power-saving mode, and leverages an advanced process node developed in partnership with TSMC. With sub-10J/TH already realized at the chip level, it’s only a matter of time before ASIC hardware can consistently deliver this level of efficiency in mining racks.
The following content is exclusively for our Premium Members:
Optimizing Efficiency with Custom Firmware
Normal Efficiency Mode (NEM) & High Efficiency Mode (HEM)
Importance of Management and Monitoring Software
Optimizing for Efficiency Doesn’t Always Lead to the Quickest ROI