DISRUPTIVE TECH
AI training is the most technologically challenging part of the overall AI process and if the network is a bottleneck , costly processing time is wasted . To speed up training , GPUs need to be interconnected in a high-performance structure known as the back-end fabric , which supports both training and storage , explains Refat Al Karmi at Juniper Networks .
Building a high performance data network for AI workloads
76 INTELLIGENTCIO MIDDLE EAST www . intelligentcio . com