Input Stories

 
Increased artificial intelligence use will put extra stress on U.S. networking capabilities (The Wall Street Journal, subscription). 
 
What’s going on: “[M]ore people will use AI chatbots and agents, which will talk in turn to still more AI agents—and all that requires more data, computing and back-end technology systems like networking … [which] is considered the ‘plumbing’ that moves data and applications inside and between data centers, as well as between data centers and internet-connected devices.” 

• The resulting amount of traffic will be “monumental,” one chief technology officer told the Journal.
• Chipmakers, networking equipment manufacturers, internet carriers and data center providers “are eyeing opportunities in a network revamp, which could include gear upgrades, new software tools and working with network providers to increase capacity and capability.”  

Why it’s happening: “Training and using generative AI models requires more data movement—known as high bandwidth—at faster speeds—known as lower latency—than other types of internet traffic.”

• What’s more, network upgrades are required to support graphics processing units, which are used to train and run AI—and costs are driven higher by the fact that “AI-ready switches cost at least five times more than traditional data center switches.”  

Cloud preferred: Many companies are training AI models in the cloud instead of in their data centers, putting pressure on cloud providers to up their network capacity.

• Amazon, for example, has invested in building its “own data centers with GPUs and AI-ready networking.”