A study by Stanford’s Hazy Research group reveals that local AI efficiency increased by 18 in 16 months, thanks to improvements in architecture and hardware. The new ‘July intelligence’ metric and hybrid models promise to reduce cloud inference costs by up to 80%. *** The research presents IPJ and IPW metrics to measure the efficiency of local AI, and finds an 18x improvement in IPJ between 2024 and 2025. Advances in hardware (such as NVIDIA B20…
This story is only covered by news sources that have yet to be evaluated by the independent media monitoring agencies we use to assess the quality and reliability of news outlets on our platform. Learn more here.
A study by Stanford’s Hazy Research group reveals that local AI efficiency increased by 18 in 16 months, thanks to improvements in architecture and hardware. The new ‘July intelligence’ metric and hybrid models promise to reduce cloud inference costs by up to 80%. *** The research presents IPJ and IPW metrics to measure the efficiency of local AI, and finds an 18x improvement in IPJ between 2024 and 2025. Advances in hardware (such as NVIDIA B20…