Heat: Satellite’s meat is GPU’s poison
Zhehu Yuan, Jinyang Liu, Guanqun Song, Ting Zhu · 2025
Read Heat: Satellite’s meat is GPU’s poison by Zhehu Yuan, Jinyang Liu, Guanqun Song, Ting Zhu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-14757.txt.
Introduction
Satellites swing between sunlight and shadow, freezing in eclipse, while their graphics processors need steady warmth to work well yet also shed heat. Zhehu Yuan and three co-authors explore a swap in a 2024 paper. The processors' own waste heat could warm the satellite during eclipse, replacing dedicated heaters. At the same time, the cold of space cools the chips.
They measure how the chips heat under different workloads, heavy in instructions versus heavy in arithmetic. The authors weigh cost savings against operating limits and design complexity, and sketch how smarter scheduling and testing on real missions could make processor heating practical for low-orbit fleets.
Catalog record
| CREATOR | Zhehu Yuan, Jinyang Liu, Guanqun Song, Ting Zhu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 24 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-14757
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Heat: Satellite’s meat is GPU’s poison Zhehu Yuan, Jinyang Liu, Guanqun Song, Ting Zhu Abstract In satellite applications, managing thermal conditions is a significant challenge due to the extreme fluctuations in temperature during orbital cycles. One of the solutions is to heat the satellite when it is not exposed to sunlight, which could protect the satellites from extremely low temperatures. However, heat dissipation is necessary for Graphics Processing Units (GPUs) to operate properly and efficiently. In this way, this paper investigates the use of GPU as a means of passive heating in low-earth orbit (LEO) satellites. Our approach uses GPUs to generate heat during the eclipse phase of satellite orbits, substituting traditional heating systems, while the GPUs are also cooled down during this process. The results highlight the potential advantages and limitations of this method, including the cost implications, operational restrictions, and the technical complexity involved. Also, this paper explores the thermal behavior of GPUs under different computational loads, specifically focusing on execution-dominated and FLOP-dominated workloads. Moreover, this paper discusses future directions for improving GPU-based heating solutions, including further cost analysis, system optimization, and practical testing in real satellite missions. I Introduction Thermal management is one of the most critical challenges for satellite operations in space, especially for small satellites that face extreme temperature variations between the sunlit and eclipse phases of their orbits. For satellites, temperature fluctuations can reach 300 degrees Celsius in a few hours [ 1 ] . Such a large temperature difference in a short period of time could cause serious results, including thermal fatigue [ 2 , 3 ] , disable of electric parts [ 4 ] , accelerate battery wear and tear [ 5 , 6 , 7 , 8 ] , etc, which will finally reduce the life span of satellites. One of the solutions is using passive thermal control systems [ 9 , 10 ] , including but not limited on thermal straps, passive radiator, and heat pipes. These methods could passively maintain the temperature of satellites within a suitable range. However, these designs are delicate and vulnerable [ 11 , 12 ] . If any part of the system are broken, the whole system may be failed. Also, these systems cannot be changed after the satellite is launched [ 13 ] . If something unexpected happens outside the design, people have no way to modify the system accordingly. In this way, the passive thermal control systems are not reliable. Moreover, their thermal control ability is relatively weak. For those extreme thermal control requirements, passive thermal control systems have nothing to do with them [ 14 , 15 , 16 ] . To mitigate these limitations of passive thermal control systems, active thermal control methods are designed [ 17 , 18 , 19 , 20 , 21 ] , such as heaters and coolers. These devices could actively control the temperature of the satellites, which make them robust against partial failures and flexible to be adapted for out-of-design situations. When some of them are disabled, the others could increase their power to produce more heat. When out-of-design situations happen, people can manually control the thermal controller to adapt to the situations. However, their prices present significant challenges, particularly for smaller satellites with limited budget [ 22 ] . In addition, these active temperature control devices require energy to operate, leading to inefficiencies and energy waste. Heat is the meat of satellites, but it is also the poison of GPU. During the computation of GPU, a large amount of waste heat are generate. If heat dissipation is not managed promptly, the GPU’s efficiency will decline, and it may even become inoperative [ 23 , 24 , 25 ] . In this study, we propose a novel approach to thermal control of satellites that utilizes GPUs as a heater in satellites. In one way, satellites are benefited with the efficiency of GPU in producing heat. On the other hand, the power consumed on heating is not wasted, while they are also used to do computation. Even better, while heating the satellites, GPUs are also cooled down, which helps them to maintain a better efficiency. In this paper, we made the following contributions: • We proposed a thermal control system that using GPU as a heater. • We compared the prices, sizes, and power of existed representative GPUs and satellite heaters. • We proposed a way to control the heat produced by GPU through job assignment. • We measured the heat produced by GPU running different jobs. II Background II-A GPU Side When GPUs are running, they produce a significant amount of energy in the form of heat because almost all the electrical energy they consume is converted into heat. For example, the NVIDIA GTX 1080 Ti, a high-end model released in 2017, has a Thermal Design Power (TDP) of 250 watts. This means that it can consume up to 250 watts of electrical power under full load, and almost all of this power is dissipated as heat [ 26 ] . Temperature plays a critical role in optimizing GPU power efficiency. The subthreshold leakage current, which is a significant contributor to power consumption, is proportional to T^{2}e^{-b/T} , where T is the temperature. This relationship shows that when the temperature increases, the leakage current increases, leading to higher power consumption. In contrast, maintaining lower temperatures reduces leakage current and improves GPU power efficiency. Therefore, lowering the temperature is essential to achieve optimal performance per watt in GPUs [ 27 ] . II-B Satellite Side Satellites experience cyclic extreme thermal conditions due to the alternating sunlit and eclipse zones, significantly affecting operational efficiency. For example, deployable solar panels (DSPs) are particularly susceptible to temperature fluctuations caused by the space environment. Maintaining the operating temperature range of solar cells ( -40^{\circ}C to +125^{\circ}C ) is crucial, as inadequate thermal management can lead to reduced efficiency or functional failure [ 28 ] . Small satellites face unique thermal challenges due to their small heat capacity, compact designs, limited radiator areas, and power constraints, making thermal control even more complex than for larger satellites [ 29 ] . For example, a small satellite with an internal heat load of approximately 9.2 W may require thermal systems that account for variations of up to 13 W to effectively manage uncertainties [ 29 ] . To address these challenges, thermal control methods are created to ensure that satellite components operate within their designated temperature ranges in an extreme environment. An approach is a passive thermal control system (TCS), as demonstrated by Budiantoro et al. (2022), which uses surface coatings such as black anodized and SG121FD white paint on DSP to regulate heat absorption and radiation [ 28 ] . This method is power efficient and cost effective, but is entirely based on material properties, making it less adaptable to rapidly changing thermal conditions or extreme temperature fluctuations, which may impact system performance. Another passive method involves the use of heat pipes for localized cooling, as explored by Elshaer [ 30 ] . Heat pipes efficiently transfer excess heat to radiative surfaces without consuming power during operation, addressing high power heat dissipation challenges [ 30 ] . However, they are limited