IBM's energy conservation projects share worldwide 2018-2022, by type

In 2022, almost two thirds of IBM's energy conservation savings were attributed to IT server efficiency projects, while 30 percent were attributed to heating and cooling related projects. Between 1990 and 2020, IBM conserved 9.9 million MWh of electricity, and was able to avoid 4.6 million metric tons of CO₂ emissions.

Data center energy efficiency

Automated, analytics-based systems are central in efforts to drive greater data center efficiency. Cooling in a data center involves separating hot air from cold air; too much hot air circulating in a data center can cause equipment to overheat and malfunction. The power usage effectiveness ratio, or PUE, is a measure of a data center’s energy efficiency, dividing total energy used by the data center by the energy consumed by the IT equipment. In an ideal world, a PUE of 1 is hoped for, where 100 percent of electricity consumption goes toward useful computation. Cooling, along with power supplies, lighting, and other factors, uses power but doesn’t compute, and as a result raises a data center’s PUE.

Future challenges

Artificial Intelligence (AI), high performance computing (HPC), and the Internet of Things (IoT) all demand faster and complex computing services. The rapidly growing cloud services segment, along with the growth of edge computing is creating ever growing energy demands and a need for efficient use of data center space. Data center providers are driven to demand more computing cores per square foot, maximizing the computational output their facilities can provide. One solution involves utilizing the power of graphics processing units (GPUs); data centers that support GPU servers can provide much higher efficiency and performance, use less power for advanced workloads, and require less floor space.