To warm your toes, most space heaters utilize high-resistance wires to generate heat. In contrast, the Heatbit small appliance heats your living space by engaging in cryptocurrency mining, developing AI for large corporations, or training advanced language models. While in operation, it rewards users with up to $28 monthly through this process.
During my encounter with CEO Alex Busarov at the Web Summit in Lisbon, he expressed, “Our primary focus is on zero-energy computing.”
Busarov highlights that the collective energy consumption from Bitcoin mining and AI training, which has surged significantly in the past seven years, now surpasses that of the entire United Kingdom. This excessive energy is essentially dissipated as heat, often necessitating additional cooling in data centers, further escalating the energy demands.
Heatbit’s innovative approach involves distributing AI processing tasks across numerous households. By repurposing the otherwise wasted heat as a valuable byproduct to warm spaces and leveraging this energy to perform essential computations, the company aims to optimize energy utilization. According to their estimates, every kilowatt-hour utilized for computing in a traditional data center incurs an additional 0.5 kWh for cooling. When individuals utilize an extra 1 kWh for heating their homes or offices, the total energy consumption amounts to 2.5 kWh between the server and the dwelling.
The solution proposed by Heatbit for both room heating and AI training involves merely utilizing 1 watt in a household. By harnessing the heat directly, bypassing the need for cooling, and eliminating the data center intermediary, a substantial 60% reduction in energy consumption is achieved. This approach underscores a greener and more efficient computing model.
Busarov asserts, “We make the same energy work twice for you.”
In regions like British Columbia, Canada, where hydroelectric power is cost-effective and eco-friendly, operating the Heatbit device would be nearly cost-free. Despite an estimated \(32 operational cost, the incentives amount to a \)28 rebate. Conversely, at standard U.S. energy rates, the operational cost would be \(40, offset by the same \)28 rebate.
To handle computational tasks and provide heating, the system incorporates 55 5-millimeter chips with 20 billion transistors. It is equipped with a money-back guarantee, can be managed through a mobile application or an on-device touchscreen, and features an HEPA filter for heating and air purification. Heatbit claims to effectively heat up to 400 square feet while operating at under 40 degrees of noise, ensuring a quieter experience than a typical refrigerator.
Priced between \(800 and \)1,000, the Heatbit device offers a comprehensive solution albeit at a relatively higher cost.
With a monthly saving of $28, the device would yield a minimum payback period of 28 months, considering it may not be as essential during the summer months (although it can still purify the air). Nonetheless, users would still be accountable for the power costs, which might be offset by the efficiency of Heatbit compared to traditional space heaters.
Initially launched through a crowdfunding campaign focused on cryptocurrency, Heatbit garnered over $1 million in unit sales. The latest iteration emphasizes AI training applications.
The challenge of managing excess heat during summer months is addressed by Busarov’s vision to distribute Heatbit devices globally, ensuring a continuous synergy between cooling and heating processes to support AI training initiatives.
