Fig. 2 shows that improved PUE and reduced electrical carbon intensity account for only small improvements when applied to the maximal IT-efficiency case. These results indicate that characteristics implying IT-device efficiency including the presence of efficient-rated devices, high-capacity utilization, IT power management and virtualization — are probably the best proxy metrics for low-carbon data centres in the near term
Figure 2 also suggests that PUE — despite its status as the de facto data centre energy-efficiency metric — is by itself a suboptimal proxy for both absolute energy and carbon performance.
Low-carbon electricity without low-energy operations is suboptimal because, given their high energy intensity, data centres now using renewable power almost exclusively draw from centralized and resource-constrained supplies (notably hydropower, wind, geothermal and biogas).
Our results and previous analyses suggest that IT-device efficiency is the most important characteristic to be reinforced through lowcarbon data centre incentives5,6,44–45. Although renewable energy can significantly reduce a data centre’s carbon emissions, an inefficient (that is, high-energy) data centre will use far more low-carbon electricity than is technically required.
→ PUE is not a good metric for absolute energy and carbon performance. If the IT device efficiency is maximized, a poor PUE (e.g. 1.8) is not that important. IT efficiency is much more important.
Grafana helped the company reduce its data center’s electricity usage by 15%.
With Europe facing an energy crisis, Sentry Software decided to leverage Grafana and OpenTelemetry to analyze its energy footprint and identify opportunities for optimization. The team was able to increase its server room’s temperature from 18°C to 27°C while still maintaining optimal conditions, resulting in significant energy savings.
