Data centers: the blind spot of GreenOps

The promise of the cloud is lower infrastructure costs compared with an on premise IT infrastructure. However, in practice, many companies use more resources than they actually need, leading to significant cost overruns.

In this context, the FinOps approach involves monitoring and optimizing cloud computing costs, so that cloud usage matches needs and cost aligns with the value delivered. This approach must be adopted both during the Build phase, to control the costs of new bills, and in Run, to continuously optimize costs and prevent drift over time.

The GreenOps concept is an extension of the FinOps approach. Its objective is to identify unused or underused IT resources, in order to optimize their use and reduce the cloud’s environmental footprint. Some companies have broadened the approach: closely linked to the worlds of FinOps or ITAM (IT asset management), GreenOps is gradually becoming an operating model that encompasses all efforts to control costs while reducing the environmental impact of all the organization’s IT infrastructures, whether hosted in-house or in the cloud. According to analyst firm Gartner, 50% of CIOs should also have performance indicators linked to the sustainability of their IT by 2025.

GreenOps is a recent concept, so it is not surprising that its scope is still variable, with no consensus yet reached. In most definitions of GreenOps, the focus is on the use phases of hardware, and therefore on their energy consumption and the consumables used. Yet while these data are of course fundamental, this approach is entirely incomplete for calculating the environmental impact of digital technology.

It is essential to take into account the entire life cycle of hardware, from the extraction of the raw materials needed to manufacture IT equipment through to its recycling. Thus, alongside consumption indicators during the use phase of IT equipment, indicators for the manufacturing phase (raw material extraction and transport), the transport phase (to the end user), and the end-of-life phase (processing, recycling, etc.) must be considered in order to effectively assess the true environmental burden of the digital value chain.

While many disparities appear among the multiple definitions of GreenOps, they all have one thing in common: the systematic absence of consideration for the entire life cycle of the data center itself, as a building, and of the equipment (IT, cooling, etc.) it houses. Yet its location, design, and layout have a significant environmental impact.

In other words, beyond the contributions of GreenOps, it now seems essential to adopt a more global sustainable IT approach. According to the Institut du Numérique Responsable (INR), this is a continuous improvement approach that aims to reduce the environmental, social, and economic impact of digital technology.

Applied to the data center, it provides companies with a methodology and indicators to reduce the environmental footprint of their data center construction, operations, maintenance, and end-of-life projects: reducing or avoiding environmental impacts from the site selection stage, conducting a data center environmental impact assessment from the design phase, reusing premises previously dedicated to other uses, low-carbon-footprint certified equipment, etc.

In all cases, this means adopting reference methodological frameworks such as the environmental assessment framework for hosting services in data centers and cloud services and standardized indicators (PUE, WUE, REF, etc.) to implement a sustainable IT approach. And above all, not limiting the approach to the carbon indicator alone—useful but insufficient to reflect, on its own, the full range of digital technology’s environmental impacts.

By Georges Ouffoué, Head of the Innovation Division, APL Data Center.