The explosion in use of AI tools across the world is increasing exponentially, but the environmental impact isn’t expressed in detail often by the companies that make these tools.
But Google has just released a technical paper detailing measurements for energy, emissions and water use of its Gemini AI prompts — and the impact of a single prompt is, it says, minuscule. According to its methodology for measuring AI’s impact, a single prompt’s energy consumption is about the equivalent of watching TV for less than 9 seconds.
That is quite low, but consider the variety of chatbots that are used, and that billions of prompts are easily sent every day.
The good news is that the technology behind these prompts has become more efficient in the past 12 months. Google says that the energy of a single Gemini text prompt has reduced by 33x and total carbon footprint has reduced by 44x. That’s not unsubstantial, and that type of momentum will need to be maintained going forward, Google says.
Google did not immediately respond to CNET’s request for further comment.
## Google’s calculation method considers much more
The search giant says the typical calculation for the energy cost of an AI prompt ends at the active machine it’s been run on, which shows a much smaller per-prompt footprint. But Google’s method for measuring the impact of a prompt spans a much wider range of factors that paint a clearer picture, including full system dynamic power, idle machines, data center overhead, water consumption and more.
For comparison, it’s estimated that only using only the active TPU and GPU consumption, a single Gemini prompt uses 0.10Wh of energy, 0.12mL of water and emits 0.02 gCO2e. This is a minute and promising number, but Google’s wider methodology tells a different story. With more considerations in place, a Gemini text prompt uses 0.24Wh of energy, 0.26mL of water and emits 0.03 gCO2e — almost more than double across the board.
## Will new efficiencies keep up with AI use?
Through a multi-layered series of efficiencies, Google is continually working on ways to make AI’s impact grow smaller. From more efficient model architectures and data centers to custom hardware, Google’s approach to addressing AI’s impact on the world is a full-stack one.
With smarter models, use cases and tools coming out by the day, those efficiencies will be much needed everywhere as we are steeped further into our new AI reality.
For more, you should stop using ChatGPT for these things.
But Google has just released a technical paper detailing measurements for energy, emissions and water use of its Gemini AI prompts — and the impact of a single prompt is, it says, minuscule.
According to its methodology for measuring AI’s impact, a single prompt’s energy consumption is about the equivalent of watching TV for less than 9 seconds.
Google says that the energy of a single Gemini text prompt has reduced by 33x and total carbon footprint has reduced by 44x.
For comparison, it’s estimated that only using only the active TPU and GPU consumption, a single Gemini prompt uses 0.10Wh of energy, 0.12mL of water and emits 0.02 gCO2e.
With more considerations in place, a Gemini text prompt uses 0.24Wh of energy, 0.26mL of water and emits 0.03 gCO2e — almost more than double across the board.
The explosion in use of AI tools across the world is increasing exponentially, but the environmental impact isn’t expressed in detail often by the companies that make these tools.
But Google has just released a technical paper detailing measurements for energy, emissions and water use of its Gemini AI prompts — and the impact of a single prompt is, it says, minuscule. According to its methodology for measuring AI’s impact, a single prompt’s energy consumption is about the equivalent of watching TV for less than 9 seconds.
That is quite low, but consider the variety of chatbots that are used, and that billions of prompts are easily sent every day.
The good news is that the technology behind these prompts has become more efficient in the past 12 months. Google says that the energy of a single Gemini text prompt has reduced by 33x and total carbon footprint has reduced by 44x. That’s not unsubstantial, and that type of momentum will need to be maintained going forward, Google says.
Google did not immediately respond to CNET’s request for further comment.
Google’s calculation method considers much more
The search giant says the typical calculation for the energy cost of an AI prompt ends at the active machine it’s been run on, which shows a much smaller per-prompt footprint. But Google’s method for measuring the impact of a prompt spans a much wider range of factors that paint a clearer picture, including full system dynamic power, idle machines, data center overhead, water consumption and more.
For comparison, it’s estimated that only using only the active TPU and GPU consumption, a single Gemini prompt uses 0.10Wh of energy, 0.12mL of water and emits 0.02 gCO2e. This is a minute and promising number, but Google’s wider methodology tells a different story. With more considerations in place, a Gemini text prompt uses 0.24Wh of energy, 0.26mL of water and emits 0.03 gCO2e — almost more than double across the board.
Will new efficiencies keep up with AI use?
Through a multi-layered series of efficiencies, Google is continually working on ways to make AI’s impact grow smaller. From more efficient model architectures and data centers to custom hardware, Google’s approach to addressing AI’s impact on the world is a full-stack one.
With smarter models, use cases and tools coming out by the day, those efficiencies will be much needed everywhere as we are steeped further into our new AI reality.
For more, you should stop using ChatGPT for these things.
