It's been almost 10 years since the U.S. federal government set out to make its own data centers more efficient and less costly to run. Over the course of those years, the world of IT has undergone a lot of change, including a continued rise in cloud computing and more widespread implementations of artificial intelligence or AI. Those changes have hit government as much as they have commercial data centers.
Since the start of its data center project, the U.S. has consolidated and closed some data centers, sure. But it's also making the most of new technologies such as artificial intelligence and sensors to improve its remaining data centers. A new report from IDC looks at the state of U.S. data centers today and some of the big initiatives that are driving them into the future.
Today's government data centers are fewer, larger, and more efficient, IDC research director Shawn P. McCarthy told InformationWeek in an interview. McCarthy focuses on data center artificial intelligence and government at IDC, and he is the author of the new data center report, IDC Planscape: AI in the Government Datacenter -- Paths to Success.
"Government data centers are undergoing a once-in-a-generation level of transformation," McCarthy said in the report. "This comes at a time when smaller data centers are consolidating into larger ones as governments across the globe are learning to deal with an ever-growing deluge of data from millions of IoT devices."
For many years the government has relied on data center infrastructure management (DCIM) tools to optimize data center computing, energy consumption, bandwidth capability, and more. Now federal data centers are looking to AI for the next stage to enable digital transformation within agencies.
McCarthy said that some of the remaining government data centers are comparable to commercial data centers in terms of advanced technologies and security.
Overall, government AI and data center spending is up, with AI-controlled DCIM managing virtualization and boosting energy efficiency, security and auto-mitigation, according to the report.
"We are entering an era where AI and DCIM will become forever entwined for managing government data centers," McCarthy said. "The insights that can be gained and the potential for intelligent machine learning -- to monitor issues and auto-address problems -- means that AI for data center management will be a hot investment area for the next decade."
McCarthy said that automation and AI go hand-in-hand in government data centers. But there's a bit of a learning curve to getting the most value and efficiency out of implementing these technologies in the data center.
"You may need a systems integrator that has worked with these kinds of solutions before," he said. For instance, in an environment full of sensors with AI designed to make decisions, every place that AI touches needs to be properly integrated and configured to deliver the most value.
"It may not be worth keeping someone on staff who has all those certifications," he said. Instead, you may want to hire a systems integrator who is an expert in that niche area to get the system up and working.
McCarthy noted the case of a data center sensor that listens to the noise generated by a computer server or other equipment and then diagnoses a need for maintenance or replacement. Cybersecurity is another important area where AI can play a critical role, McCarthy said. AI can help monitor and manage this for big government data centers and adjust as changes are needed.
"AI has found a home in data center management, and as government agencies work to make their facilities more efficient, AI will continue to be a major government system management resource," McCarthy said in the report.
For more on AI and automation, read: