How DCIM Improves Data Centre Efficiency and PUE

Many data centres operate at unnecessarily cool temperatures, typically between 18°C and 20°C. This happens because of an outdated concern about hardware failing. However, today’s servers are built to work well even when the air is warmer.

• The Power of the Set Point: Data Centre Infrastructure Management (DCIM) plays a key role here. It uses sensors that constantly monitor the environment. These sensors create a detailed picture of your data centre’s “thermal envelope.” This is essentially the range of temperatures within the facility. Consequently, allowing you to adjust CRAC set points accordingly.
• The 4% Rule: The “4% Rule” shows a direct link between higher temperatures and lower energy costs. Specifically, for every 1°C you increase the air temperature, you can save about 4% on your cooling expenses. To illustrate, imagine you raise the cooling set point from 19°C to 24°C. This is a 5°C increase. Such a change could lead to a substantial 20% reduction in the energy used for cooling.
• DCIM Precision: Without DCIM, managers often cool the entire room just because one rack hits 32°C. However, DCIM helps you identify localised hot spots and cold bypasses. This allows you to fix specific airflow issues instead of lowering the temperature for the whole data hall.

Achieving optimal data centre efficiency is crucial. The industry has a standard way to measure this. It is called Power Usage Effectiveness, or PUE.

PUE is calculated by looking at two key numbers. The first is the total energy used by the entire facility. The second is the energy used only by the IT equipment inside. The formula is simple: PUE equals Total Facility Energy divided by IT Equipment Energy. A lower PUE number means better efficiency.

Traditionally, finding your PUE has been a difficult process. It often involved a lot of manual work. This work was not done very often. As a result, the PUE data you got was usually outdated, or “stale.” This made it hard to make timely improvements.

However, there’s a better way. DCIM applications can provide live PUE monitoring. This means you can see your PUE in real-time. You can watch how it changes. For example, you can see the effect of a heatwave. You can also see how PUE changes during times of heavy computer use. This constant insight is invaluable.

Furthermore, these tools offer transparent reporting. You can generate necessary reports with just a few clicks. These reports are perfect for sharing with important people. This includes stakeholders or official groups. They help prove your facility is living up to its eco-friendly goals. You can easily show your commitment to being “green.”

One of the biggest hidden expenses in data centres is stranded capacity—unused power and space that has already been allocated. This inefficiency drives up costs, but it can be fixed with better visibility.

• The Problem: Misleading Assumptions – Without Data Centre Infrastructure Management (DCIM), you might assume a rack is “full” simply because it has reached its theoretical power limit. For example, if a rack is rated for 10kW, teams often stop using it once they reach that number—even if the actual power draw is lower.
• The Solution: Real-Time Data with DCIM – DCIM tools, like Sensorium, provide real-time monitoring to show the actual power usage. If that same 10kW rack is only using 5kW, you still have 5kW of available (“free”) capacity. Instead of buying more equipment or expanding the facility, you can optimise what you already have.
• The Result: Lower Costs, Higher Efficiency – By tracking true resource usage, DCIM helps avoid costly expansions. This means getting the most value from existing space and power—improving ROI and reducing waste.

Stranded capacity is wasted money. With DCIM, you turn unused resources into opportunities for growth.

Many UK data centres house a silent, wasteful problem: “ghost” or “zombie” servers. Shockingly, up to 30% of their machines can fall into this category. These are servers that remain physically plugged in and consume electricity. However, they are no longer performing any useful computing work. They are simply drawing power for nothing.

• Detection: The solution lies in Data Centre Infrastructure Management (DCIM) systems. These systems actively scan for hardware. Specifically, they look for units with high power consumption but almost zero CPU or data activity. This contrast is a clear red flag.
• The Saving: Decommissioning these “ghosts” brings significant advantages. Firstly, it leads to an immediate and noticeable reduction in electricity bills. Furthermore, it frees up valuable rack space. This newly available space can then be reused for active, revenue-generating equipment. Therefore, proactively hunting these ghost servers directly translates into cost savings and increased operational efficiency.