IBM's claim is that the power reduction for training one AI model would be enough to run 5,000 US homes for a year and using light would reduce the time for training an AI Large Language Model ...
A big reason, say IBM Corp. researchers, is that the connections within data centers still rely largely on copper wire instead of speedy fiber optics. “We estimate GPUs spend about half their ...
This advancement could reduce GPU downtime, accelerate AI training processes, and significantly lower energy consumption, positioning IBM as a leader in the quest for more sustainable computing.
IBM said last year that it would replace 7,800 jobs with AI, or up to 30% of the company, proving those fears that AI is coming for human jobs has some truth to it. As it stands today, AI could ...
PCMag editors select and review products independently. If you buy through affiliate links, we may earn commissions, which help support our testing. IBM has developed new optics technology that ...
IBM says that with its CPO optics, AI models can be trained much faster in data centers as well, which contributes to those reduced energy costs. It can bring down AI model training time from ...