Both overclocking and autotuning involve adjusting the frequencies on the hashing boards to improve the performance of the ASIC. The difference between the two lies in the complexity of these frequency settings.

Overclocking is a simple adjustment that involves increasing the frequency on the hashing boards to increase the machine's hash rate. Underclocking, on the other hand, involves reducing the clock frequency of an integrated circuit to a level below its nominal frequency. By reducing the clock frequency of an ASIC through underclocking, the heat generated by the hardware is also reduced. This allows for a decrease in the speed of the fans required to cool the ASIC, as they do not have to work as hard to maintain an appropriate temperature. By reducing the fan speed, the noise generated by the ASIC is also reduced. This can be particularly useful for users who use ASICs at home and are looking to minimize the noise disturbances caused by mining equipment.

Braiins OS+ supports overclocking, underclocking of ASICs, and autotuning. It allows users to adjust the clock frequency of their hardware flexibly, maximizing performance or saving energy according to their preferences.

Before 2018, miners had two ways to gain an advantage in their activity: finding electricity at a reasonable cost and buying more efficient hardware. However, in 2018, a new advancement was discovered in the field of mining software and firmware, called AsicBoost. This technique allows miners to reduce their costs by approximately 13% by modifying the firmware running on their devices.

Today, a new advancement in the software and firmware mining sector has emerged, called autotuning, which offers an even greater advantage than AsicBoost. ASICs are composed of many small computer chips that perform hashing. These chips are made of silicon, the same element widely used in semiconductors and other microelectronic components. The key understanding here is that not all silicon chips are identical; each can vary slightly in its electrical properties. Hardware manufacturers are aware of this and publish the performance specifications of their mining machines based on the lower limit of their tolerances. In other words, manufacturers know the frequency that works best for average chips, and they use this frequency uniformly for all chips in the machine.

This puts an upper limit on the hash rate a machine can have. Autotuning is a process in which algorithms evaluate the optimal frequencies for chip-by-chip hashing, instead of treating the entire machine as a single unit. This means that a higher-quality chip, which can perform more hashes per second, will have a higher frequency, while a lower-quality chip, which can perform relatively fewer hashes, will have a lower frequency. Chip-level autotuning is a method for optimizing the performance of an ASIC by adjusting the software and firmware running on it.

The end result is a higher hash rate per watt of electricity, which means larger profit margins for miners. The reason machines are not distributed with this type of software is that machine variance is undesirable, as customers want to know exactly what they are getting. Therefore, it is a bad idea for manufacturers to sell a product that does not offer consistent and predictable performance across different machines. Additionally, chip-level autotuning requires considerable development resources, as it is a complex implementation. Manufacturers already spend a lot of resources developing their own firmware. There are software solutions that allow for autotuning, such as Braiins OS+. In addition to improving ASIC performance by up to 20%.