The contribution of electricity consumption of chatt – less than expected? Felix Baumann first appeared on Basic Thinking. You always stay up to date with our newsletter.
Chatgpt’s electricity consumption has triggered a heated debate, because a search query should consume ten times as much electricity as on Google. According to a current study, the previous assessments could be exaggerated.
Experts assume that an inquiry to Chatgpt uses many times more energy than a search query on Google. According to one often quoted study From 2023, Chatgpt consumes around 2.9 watt hours of electricity for a single answer, while a search query on Google only needs about 0.3 watt hours. Extrapolated to one year, this would result in an energy consumption of 227 million kilowatt hours. That would be enough to supply small states like Gibraltar for a year.
One new investigation However, the Epoch AI research institute shows that these values may be very exaggerated. The researchers carried out current measurements and came to the conclusion that a standard request to Chatgpt also “only” used 0.3 watt hours-exactly as much as a Google search. The drastic deviation to the earlier estimate could be explained by more efficient hardware and optimized models.
Chatgpt: Electricity consumption depends on interaction
Openaai has made considerable progress in energy efficiency since 2023, among other things through improved data centers and optimized use of tokens, the smallest processing units within the AI. However, the actual electricity consumption depends heavily on how users interact with chatt.
While simple text inquiries are in the range of 0.3 watt hours, longer or more complex inputs can quickly need up to 2.5 watt hours according to the study. In extreme cases, for example with elaborate calculations or the processing of gigantic data, consumption for an inquiry is up to 40 kilowatt hours. This corresponds to the double power consumption of an average US budget per minute.
Search engines are increasingly relying on AI
Nevertheless, the new study shows that the energy consumption of AI applications should be viewed more differently. Although powerful AI models such as Chatgpt are energy-intensive, modern optimizations ensure that they do not necessarily consume more electricity than conventional search engines.
Since Google is also increasingly using AI-based search technologies, the energy requirement should continue to approach in the future. However, the discussion about the electricity consumption of artificial intelligence remains up to date. While chatt and similar models need enormous computing power, developers continuously work on making technology more efficient. In the long term, the electricity requirement could drop, while AI applications become more efficient at the same time.
Also interesting:
- Artificial intelligence: What is a prompt?
- Robot recognize human touch – without artificial skin
- Artificial intelligence in the iPhone 16: These are the new Apple products
- Self -healing power grid: Artificial intelligence should avoid blackouts
The contribution of electricity consumption of chatt – less than expected? Felix Baumann first appeared on Basic Thinking. Follow us too Google News and Flipboard.
As a Tech Industry expert, I am impressed by the efficiency of ChatGPT’s power consumption. Given the complexity of natural language processing and the immense computational power required for such tasks, the fact that ChatGPT is able to operate with less power than expected is truly remarkable.
This not only demonstrates the advancements in AI technology but also shows a commitment to sustainability and reducing environmental impact. With the increasing focus on green technology and energy efficiency, ChatGPT sets a positive example for other AI models and technologies in the industry.
Overall, the lower-than-expected power consumption of ChatGPT is a promising development that highlights the potential for more sustainable and efficient AI solutions in the future.
Credits