To mitigate the growing carbon footprint of computing systems, there has been an increasing focus on carbon-aware approaches that seek to align the power usage of IT infrastructure with the availability of clean energy. Unfortunately, research on carbon-aware applications and the required interfaces between computing and energy systems remain complex, due to the scarcity of available testing environments. To this day, almost all new approaches are evaluated on self-implemented simulation testbeds, which leads to repeated development efforts by researchers and low comparability of approaches. In this paper, we present our vision of a co-simulation testbed for carbon-aware applications and systems. We envision a versatile testbed which lets users connect domain-specific simulators for components like renewable power generation, energy storage, and power flow analysis with real software and hardware. By providing extensibility on the one hand and access to state-of-the-art implementations, datasets, and best practices on the other, we hope to accelerate research in carbon-aware computing. In addition, a co-simulation testbed can be useful for development and operations, like in continuous testing. We implemented a first prototype of our idea and welcome the community to contribute to this vision.

翻译：为了缓解计算系统日益增长的碳足迹，学术界对碳感知方法日益关注，这类方法旨在使IT基础设施的能耗与清洁能源的可用性保持一致。然而，由于可用测试环境的稀缺，对碳感知应用以及计算与能源系统之间所需接口的研究仍面临复杂性。迄今为止，几乎所有新方法均在自研仿真测试平台上进行评估，这导致研究人员重复开发、方法间可比性较低。本文提出了一种面向碳感知应用与系统的协同仿真测试平台愿景。我们设想构建一个通用测试平台，允许用户将可再生能源发电、储能、潮流分析等领域的专用仿真器与实际软件和硬件相连接。通过一方面提供可扩展性，另一方面提供最先进的实现方案、数据集和最佳实践，我们希望加速碳感知计算领域的研究。此外，协同仿真测试平台在开发运维（如持续测试）中也具有实用价值。我们已实现该想法的首个原型，并欢迎社区为此愿景贡献力量。