Data Centers

Tech emissions are growing at an estimated 6% year on year, and globally, data centers now consume around 3% of the world’s electricity and produce about 2% of total GHG emissions. Industry analysts predict that these emissions will, unless addressed, rise to 10% of global emissions within a decade.

 

In recent years, there have been substantial efficiency improvements in data centers. Yet, at the same time, the industry faces ever-increasing demand for the computing power it provides – making the journey to net zero significantly challenging.

 

  • As data volumes continue to grow 30% per year, the Paris Agreement targets for the data centre industry still require significant emissions reductions of about 53% to 2030.
  • Despite substantial  investments in renewable energy, the variability of wind and solar sources may not match a data centre’s demand profile, and renewable energy power purchase agreements might even be for a different grid or region.
  • The cost of carbon offsetting is rising, and whilst renewables have their place, consumption reduction and optimisation should be the priority.
  • It is hard to find sustainability wins within existing infrastructure and reduce energy wastage throughout.

The complexity of digital has presented challenges to carbon accounting, but the availability of data also presents opportunities that are not available to more traditional industries.

 

QiO’s Foresight Optima helps data centre operators on their sustainability journey with continuous efficiency improvements in energy and water consumption at multiple levels.

Continuously optmise HVAC and cooling sytems contol, reducing energy and water consumption and improving PUE.

Monitor patterns in server power usage to automatically adjust power consumption to better match workloads.

Run optimisation in a fully autonomous closed-loop manner, with minimal interventions required to keep efficiency high.

Features

Foresight Optima provides a suite of ready-to-deploy features with integrated algorithms that integrate with existing data centre infrastructure management (DCIM) systems to recommend and implement optimal control settings on HVAC and cooling systems as well as the CPU processors themselves, helping reduce energy and carbon emissions.

Energy Efficiency Index

See opportunities to reduce energy and water consumption and costs in real-time without spending hours analysing time-series usage and workload data. The EEI algorithm recommends fan speed, temperature and even processor power state settings based on real-time server workload, resulting in quantified cost savings and reduced carbon emissions.

Energy Mix Optimisation

Use location-specific weather forecast data to accurately predict energy generation from on-site solar or wind distributed energy sources, reducing fuel consumption and grid demand during peak tariff times, resulting in reduced emissions and lower running costs.

Autonomous Control

Automatically adjust performance and sleep states to match time-of-day workloads. Run your system in open or closed-loop, either manually suggesting state changes or implementing them automatically.

Case Study

A large data centre operator company used Foresight Optima DC+ to reduce server power consumptionmnduring idle time while maintaining service quality (QoS).

The Foresight Efficiency Index used granular telemetry data to identify opportunities for power savings and suggested actions to be taken on the platform to realise estimated power savings. The system ran in closed-loop, automatically implementing the AI-suggested actions and placing specific processors in the recommended power-saving state. This resulted in reduced CPU power consumption while maintaining application QoS results, leading to a reduced carbon footprint.

DC2

Large data centre operator