Advanced Computing in the Age of AI | Thursday, March 28, 2024

Cloud Spending Grows Despite Unused Capacity 

AI and machine learning workloads along with Internet of Things and edge computing initiatives are prompting companies to add workload processing capacity. Still, according to a new survey, that new capacity is offset by the fact that new servers and other IT infrastructure often remain idle for long stretches.

The survey of workload processing demand released this week by platform vendor Kazuhm found that 70 percent of respondents plan to purchase new servers, mostly to increase workload processing capacity. (The percentage for equipment like laptops and desktops was even higher.)

At the same time, 70 percent of those surveyed said servers and other equipment are shut down for long stretches, with more than 70 percent saying servers are idle 25 percent of the time on nights and weekends.

“Workload containerization combined with the recapture of underused resources offers a secure, performant and cost-effective alternative,” the survey found.

Nevertheless, data-heavy applications will continue to drive demand for computing resources, especially agile microservices that promise to boost infrastructure utilization. For example, three-quarters of the respondents to the Kazuhm poll said they expect to increase their use of application containers this year. Meanwhile, 86 percent of those surveyed expect to boost utilization of public clouds—that despite the fact that only 42 percent of those polled consider their data to be secure in the cloud.

San Diego-based Kazuhm, which specializes in helping customers squeeze more computing power out of existing IT resources, argues that all that unused capacity can be tapped for AI/ML and IoT/edge projects without investing more in public cloud access.

“While many prepare to increase their IT spending in the cloud, concerns remain regarding data security,” the survey concludes. “Given the levels of IT resources, including desktops and servers that have unused [or] underused capacity, IT professionals have the opportunity to build a thoughtful strategy around the rollout of data-heavy applications where they choose the right resource for the right reason.”

Kazuhm said this week it polled 540 IT managers at the end of December 2018. Those surveyed included CIOs, CTOs, IT system administrators and purchasing managers.

About the author: George Leopold

George Leopold has written about science and technology for more than 30 years, focusing on electronics and aerospace technology. He previously served as executive editor of Electronic Engineering Times. Leopold is the author of "Calculated Risk: The Supersonic Life and Times of Gus Grissom" (Purdue University Press, 2016).

EnterpriseAI