AI companies face high computational costs and energy demands while developing and deploying models. Startup crops are building solutions to make AI more energy and cost-effective. .
Deepseek has sent a clear message to the Silicon Valley startup.
Now, US competitors are rushing to replicate the Chinese startup approach. This appears to be comparable to the performance of top American AI models, but it appears to be almost a fraction of the cost.
This has led industry insiders to question the billions of dollars spent on AI infrastructure. It also highlights startups developing solutions that reduce the high costs of developing, deploying and implementing AI models.
Training AI requires enormous processing power, fueled by a graphics processing unit or cluster of GPUs. They consume a lot of energy and these power circuits are primarily provided by data centers.
“The energy consumption of large-scale AI model training can generate emissions equivalent to the lifetime emissions of multiple vehicles,” Andreas Riegler, general partner at Apex Ventures, told Business Insider . “As the model grows in size, the demand for energy scales will expand exponentially, raising sustainability concerns for future applications,” he added.
Riegler has many approaches to making AI more environmentally friendly and cheaper to use. He said startups can improve software efficiency, develop more energy-efficient chips, and utilize renewable energy sources.
BI spoke to seven European and US-based investors and asked them to suggest startups that will help make AI cheaper and greener.
