SambaNova research finds that many companies don’t understand the energy demands of AI Nearly three-quarters are aware of the dramatic energy demands required to train AI models AI Energy Only 13% actively monitor consumption, which may indicate that most off-site facilities are used
The introduction of agent AI, known for its advanced decision-making abilities, is raising concerns about energy consumption, a new study claims, as energy demand increases significantly due to the move from simple algorithms to sophisticated models. .
A study by SambaNova Systems that sampled more than 2,000 business leaders in the U.S. and Europe found that 70% of business leaders recognize significant energy requirements for training models in AI tools, but do not monitor the power consumption of AI systems. We found that only 13% do. .
At the same time, 37.2% of companies are facing increased pressure from stakeholders to improve energy efficiency, and 42% expect these demands to increase further.
AI energy demand challenges
Rising energy costs have become a significant challenge, with 20.3% of businesses identifying it as a pressing issue.
Thankfully, 77.4% of businesses are actively looking for ways to reduce their power usage by optimizing models, deploying energy-efficient hardware, and investing in renewable energy solutions.
However, these efforts have not kept pace with the rapid expansion of AI systems, leaving many companies vulnerable to rising costs and sustainability pressures.
“This study reveals a harsh reality: Companies are rushing to adopt AI, but are unprepared to manage its energy impact,” said Rodrigo Liang, CEO of SambaNova Systems. says.
“Without an aggressive approach to increasing the efficiency of AI hardware and energy consumption, we risk undermining the very advances that AI promises, especially in the face of increasing demands from AI workflows.” he added.
“By 2027, we predict that more than 90% of leaders will be concerned about the power demands of AI.As enterprises integrate AI, addressing energy efficiency and infrastructure readiness will help ensure long-term success. will be essential.”
