PLUS: Microsoft to invest £2.5B in UK datacenters to power AI, and more
AI in brief Using a text-to-image model to craft an AI-generated image can require almost the same amount of power as that required to charge a smartphone, according to recent research.
In a paper released on arXiv last week, a team of researchers from Hugging Face and Carnegie Mellon University calculated the amount of power AI systems use when asked to perform different tasks.
After asking AIs to perform 1,000 inferences for each task, the researchers found text-based AI tasks are more energy-efficient than jobs involving images.
Text generation consumed 0.042kWh and generation needed 1.35kWh. The boffins assert that charging a smartphone requires 0.012kWh – making image generation a very power-hungry application.
« The least efficient image generation model uses as much energy as 950 smartphone charges (11.49kWh), or nearly one charge per image generation, » the authors wrote, noting the « large variation between image generation models, depending on the size of image that they generate. »
The authors also measured carbon dioxide created by different AI workloads. As depicted in the graphic below, image creation topped that chartMicrosoft invests £2.5B in UK datacenters
Microsoft will spend £2.5 billion to expand its datacenters in the UK and plans to fill them with over 20,000 GPUs by 2026.
The deal is reportedly the single largest investment the tech behemoth has made in Blighty since it launched operations on that side of the pond 40 years ago. Microsoft currently has datacenters in London, Cardiff, and may support other sites in northern England in the future.
Microsoft also pledged to train a million people to work in AI with a multi-million-pound investment. It will provide support for new programs to teach workers skills to help them understand applications and build the technology and plans to establish the first Professional Certificate on Generative AI.
