\n<\/p>\n
I<\/span>n mid-2023, if a user asked OpenAI\u2019s ChatGPT for a recipe for artichoke pasta or instructions on how to make a ritual offering to the ancient Canaanite deity Moloch, its response might have taken \u2013 very roughly \u2013 2 watt-hours, or about as much electricity as an incandescent bulb consumes in 2 minutes.<\/p>\n OpenAI released a model on Thursday that will underpin the popular chatbot \u2013 GPT-5. Ask that version of the AI for an artichoke recipe, and the same amount of pasta-related text could take several times \u2013 even 20 times \u2013 that amount of energy, experts say.<\/p>\n As it rolled out GPT-5, the company highlighted the model\u2019s breakthrough capabilities: its ability to create websites, answer PhD-level science questions, and reason through difficult problems.<\/p>\n But experts who have spent the past years working to benchmark the energy and resource usage of AI models say those new powers come at a cost: a response from GPT-5 may take a significantly larger amount of energy than a response from previous versions of ChatGPT.<\/p>\n OpenAI, like most of its competitors, has released no official information on the power usage of its models since GPT-3, which came out in 2020. Sam Altman, its CEO, tossed out some numbers on ChatGPT\u2019s resource consumption on his blog this June. However, these figures, 0.34 watt-hours and 0.000085 gallons of water per query, do not refer to a specific model and have no supporting documentation.<\/p>\n \u201cA more complex model like GPT-5 consumes more power both during training and during inference. It\u2019s also targeted at long thinking \u2026 I can safely say that it\u2019s going to consume a lot more power than GPT-4,\u201d said Rakesh Kumar, a professor at the University of Illinois, currently working on the energy consumption of computation and AI models.<\/p>\n The day GPT-5 was released, researchers at the University of Rhode Island\u2019s AI lab found that the model can use up to 40 watt-hours of electricity to generate a medium-length response of about 1,000 tokens, which are the building blocks of text for an AI model and are approximately equivalent to words.<\/p>\n A dashboard they put up on Friday indicates GPT-5\u2019s average energy consumption for a medium-length response is just over 18 watt-hours, a figure that is higher than all other models they benchmark except for OpenAI\u2019s o3 reasoning model, released in April, and R1, made by the Chinese AI firm Deepseek.<\/p>\n This is \u201csignificantly more energy than GPT-4o\u201d, the previous model from OpenAI, said Nidhal Jegham, a researcher in the group.<\/p>\n Eighteen watt-hours would correspond to burning that incandescent bulb for 18 minutes. Given recent reports that ChatGPT handles 2.5bn requests a day, the total consumption of GPT-5 could reach the daily electricity demand of 1.5m US homes.<\/p>\n As large as these numbers are, researchers in the field say they align with their broad expectations for GPT-5\u2019s energy consumption, given that GPT-5 is believed to be several times larger than OpenAI\u2019s previous models. OpenAI has not released the parameter counts \u2013 which determine a model\u2019s size \u2013 for any of its models since GPT-3, which had 175bn parameters.<\/p>\n A disclosure this summer from the French AI company Mistral finds a \u201cstrong correlation\u201d between a model\u2019s size and its energy consumption, based on Mistral\u2019s study of its in-house systems.<\/p>\n \u201cBased on the model size, the amount of resources [used by GPT-5] should be orders of magnitude higher than that for GPT-3,\u201d said Shaolei Ren, a professor at the University of California, Riverside who studies the resource footprint of AI.<\/p>\n GPT-4 was widely believed to be 10 times the size of GPT-3. Jegham, Kumar, Ren and others say that GPT-5 is likely to be significantly larger than GPT-4.<\/p>\n Leading AI companies like OpenAI believe that extremely large models may be necessary to achieve AGI, that is, an AI system capable of doing humans\u2019 jobs. Altman has argued strongly for this view, writing in February: \u201cIt appears that you can spend arbitrary amounts of money and get continuous and predictable gains,\u201d though he said GPT-5 did not surpass human intelligence.<\/p>\n skip past newsletter promotion<\/p>\n A weekly dive in to how technology is shaping our lives<\/p>\n Privacy Notice: <\/strong>Newsletters may contain info about charities, online ads, and content funded by outside parties. For more information see our Privacy Policy. We use Google reCaptcha to protect our website and the Google Privacy Policy and Terms of Service apply.<\/span><\/p>\n after newsletter promotion<\/p>\n In its benchmarking study in July, which looked at the power consumption, water usage and carbon emissions for Mistral\u2019s Le Chat bot, the startup found a one-to-one relationship between a model\u2019s size and its resource consumption, writing: \u201cA model 10 times bigger will generate impacts one order of magnitude larger than a smaller model for the same amount of generated tokens.\u201d<\/p>\n Jegham, Kumar and Ren said that while GPT-5\u2019s scale is significant, there are probably other factors that will come into play in determining its resource consumption. GPT-5 is deployed on more efficient hardware than some previous models. GPT-5 appears to use a \u201cmixture-of-experts\u201d architecture, which means that it is streamlined so that not all of its parameters are activated when responding to a query, a construction which will likely cut its energy consumption.<\/p>\n On the other hand, GPT-5 is also a reasoning model, and works in video and images as well as text, which likely makes its energy footprint far greater than text-only operations, both Ren and Kumar say \u2013 especially as the reasoning mode means that the model will compute for a longer time before responding to a query.<\/p>\n \u201cIf you use the reasoning mode, the amount of resources you spend for getting the same answer will likely be several times higher, five to 10,\u201d said Ren.<\/p>\n In order to calculate an AI model\u2019s resource consumption, the group at the University of Rhode Island multiplied the average time that model takes to respond to a query \u2013 be it for a pasta recipe or an offering to Moloch \u2013 by the model\u2019s average power draw during its operation.<\/p>\n Estimating a model\u2019s power draw was \u201ca lot of work\u201d, said Abdeltawab Hendawi, a professor of data science at the University of Rhode Island. The group struggled to find information on how different models are deployed within data centers. Their final paper contains estimates for which chips are used for a given model, and how different queries are parceled out between different chips in a datacenter.<\/p>\n Altman\u2019s June blog post confirmed their findings. The figure he gave for ChatGPT\u2019s energy consumption per query, 0.34 watt-hours per query, closely matches what the group found for GPT-4o.<\/p>\n Hendawi, Jegham and others in their group said that their findings underscored the need for more transparency from AI companies as they release ever-larger models.<\/p>\n \u201cIt\u2019s more critical than ever to address AI\u2019s true environmental cost,\u201d said Marwan Abdelatti, a professor at URI. \u201cWe call on OpenAI and other developers to use this moment to commit to full transparency by publicly disclosing GPT-5\u2019s environmental impact.\u201d<\/p>\n","protected":false},"excerpt":{"rendered":" In mid-2023, if a user asked OpenAI\u2019s ChatGPT for a recipe for artichoke pasta or instructions on how to make a ritual offering to the ancient Canaanite deity Moloch, its response might have taken \u2013 very roughly \u2013 2 watt-hours, or about as much electricity as an incandescent bulb consumes in 2 minutes. OpenAI released<\/p>\n","protected":false},"author":1,"featured_media":14915,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[52],"tags":[8504,611,8505,495,4112,1430],"class_list":{"0":"post-14914","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-technology","8":"tag-disclose","9":"tag-energy","10":"tag-gpt5s","11":"tag-higher","12":"tag-models","13":"tag-openai"},"_links":{"self":[{"href":"https:\/\/naijaglobalnews.org\/index.php?rest_route=\/wp\/v2\/posts\/14914","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/naijaglobalnews.org\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/naijaglobalnews.org\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/naijaglobalnews.org\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/naijaglobalnews.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=14914"}],"version-history":[{"count":0,"href":"https:\/\/naijaglobalnews.org\/index.php?rest_route=\/wp\/v2\/posts\/14914\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/naijaglobalnews.org\/index.php?rest_route=\/wp\/v2\/media\/14915"}],"wp:attachment":[{"href":"https:\/\/naijaglobalnews.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=14914"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/naijaglobalnews.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=14914"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/naijaglobalnews.org\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=14914"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}Benchmarking AI power usage<\/strong><\/h2>\n
Hidden information<\/strong><\/h2>\n