OpenAI's Inference Breakthrough Cuts Serving Costs by Over 50%
OpenAI reportedly found a big inference tuning that cuts serving costs. By more than half, according to industry reports. If true, that could sharply lower the cost of running AI models at scale. Make advanced AI easier for shops of all sizes to afford. Separately, chip startup Etched raised significant funding. It helps to build specialized inference hardware.
Inference tuning can cut serving costs by more than 50%.
Breakthrough could speed up AI adoption by lowering barriers to entry
Etched raises funding to build specialized inference chips