Dive Brief:
- Organizations are relying on third-party colocation data centers to fulfill AI-related infrastructure needs, Flexential said in its 2024 State of AI Infrastructure report released Tuesday.
- Only 24% of respondents are deploying AI hardware on-premise and 51% said they are leasing rack space in third-party colocation data centers to process data closer to the edge of their network, according to Flexential. The company polled 350 information technology decision-makers in organizations with over $100 million in annual revenue.
- IT leaders are willing to pay a premium for improved sustainability outcomes from third-party data centers or cloud vendors, with 94% of respondents saying they would pay a premium for data centers that use clean or renewable energy or buy credits to offset their carbon footprints. Forty-five percent said they were willing to pay 10% or more for sustainable cloud services, the report says.
Dive Insight:
The artificial intelligence boom has driven urgency and visibility to discussions on reducing data centers’ carbon footprints, Flexential says. By 2027, AI infrastructure could consume up to 134 TWh of power annually, the report says, citing a peer-reviewed analysis published in Joule magazine in October 2023.
Existing data center designs are often unable to meet AI’s large power demands, while the continued dominance of cloud and hyperscale computing in major markets is driving smaller enterprises to seek colocation space and power to meet their needs, according to JLL’s H2 2023 North American data center report released in February. In response to limited power availability from the public grid, data center operators are increasingly exploring microgrids that can integrate renewable energy and provide resiliency from disruptions, JLL says.
The Electric Power Research Institute in May published a report concluding data centers could consume 9% of the United States’ electricity generation by 2030, about double the amount consumed today. AI queries require about ten times the electricity of traditional internet searches, the report said.
Looking at the key elements of AI roadmaps, the report found that 59% of respondents’ organizations were increasing infrastructure investments to account for increased AI-driven workloads, and 52% saying they are working to develop in-house AI applications and solutions. An additional 52% said a focus on improving data center sustainability is a key element, while 46% said they aim to increase their use of edge computing.
These sentiments are echoed in Honeywell’s recent industrial AI report, also released Tuesday, which found overwhelming support for AI adoption. Although 37% of respondents expressed concern that their executive leadership teams do not fully understand AI, 94% said corporate leadership is “all in” on AI adoption, according to a survey of 1,600 executives in 12 global markets, Honeywell said in a news release Tuesday.
Automation is a key driver for AI implementation, according to 64% of the respondents surveyed from April 22 through May 2 by Wakefield Research, commissioned by Honeywell for the report. Sixty percent named improved cybersecurity and threat detection as a key benefit, while 59% listed the generation of real-time data to improve decision-making as an advantage of AI implementation.
AI implementation, however, can be expensive when it involves working with “an existing physical plant or technological backbone,” Honeywell warns. While 75% of AI leaders believe that their organizations will replace non-AI-compatible equipment, only 41% said they will do so early enough to maximize AI’s benefits. Thirty-three percent said they will “wait out the lifespan of legacy equipment,” according to Honeywell.
Meanwhile, organizations are increasingly drawing on AI to bridge the growing skills gap and wave of retirements. About 64% of respondents said they believe AI’s most promising use in their organizations is its ability to boost worker efficiency and productivity. Only 15% said they expect AI to cover existing labor shortages. Instead, 49% said they are using AI-enabled technology to increase worker flexibility, while 45% believe AI will help improve job satisfaction.
Forty-four percent of respondents said they would use AI to create more time for skills development. Over 80% of respondents, however, said they underestimate the training resources needed to introduce AI to their workforce, Honeywell cautioned.
