Ep 20: AI Data Centers Must Be Sustainable! with Neuralwatt CEO Chad Gibson

<p>When we make AI data centres expensive and unsustainable that's a choice because they don't have to be.</p><p></p><p>In this episode of Asynchronous and Unreliable, host Anne Currie speaks with Chad Gibson, ex Microsoft co-founder and CEO of AI host and infrastructure company Neuralwatt, to discuss innovative approaches to reducing the energy footprint of AI.</p><p></p><p>Discover how layered technologies, transparency, and customer-driven model choices can help to shape greener, cheaper, and more efficient AI.</p><p></p><p>Key topics</p><ul><li>NeuralWatt’s multi-layered approach to AI energy efficiency, from GPUs to inference cloud services</li><li>The key role of software optimization in reducing power consumption and improving model efficiency</li><li>How the demand for open source and Chinese models influences energy efficiency strategies</li><li>Challenges in serving large, long-context models like GLM 5.2 and their operational impacts</li><li>The importance of transparency and customer demand in achieving sustainable AI practices</li><li>The potential of autonomous agents for further optimizing energy consumption</li><li>Industry trends in hardware, model size, and software innovation pushing AI toward scalability</li><li>Real-world examples of data center practices that either hinder or help sustainability goals</li></ul><p></p><p><a rel="noopener noreferrer nofollow" href="https://www.asynchronousunreliable.com/episode-20-chad-gibson" target="_blank">Transcript</a></p><p><a rel="noopener noreferrer nofollow" href="https://neuralwatt.com/" target="_blank">Neuralwatt</a></p>