When it comes to determining the size of data centers, the common practice is to measure them based on their energy intake. Conversations about data center capacity often revolve around the maximum megawatts a facility can consume, leading to companies boasting about their 100 MW data center or discussing 10 MW data center projects.
But why do we use energy input as the primary metric for measuring data center size? What are the benefits and drawbacks of this approach? And what other metrics could the industry consider to compare data center capacity?
Understanding Data Center Size Measurements
When people inquire about the size of a data center, it’s usually because they want to know how many servers it can support, which ultimately determines how many applications or data it can host. This makes sense since the primary purpose of data centers is to host workloads, making it crucial to understand their hosting capacity.
However, there isn’t a single metric that can accurately capture a data center’s total hosting capacity. Various factors come into play, including:
Building size: The physical dimensions of a facility dictate how many servers can be accommodated.
Server rack size: Different-sized server racks affect the number of servers a data center can hold.
Server rack layout: The arrangement of server racks on the data center floor impacts the total number of racks and servers that can be accommodated.
Cooling capacity: The efficiency of data center cooling systems can limit the total number of servers that can be deployed.
Network infrastructure: The capacity of the network connecting servers within the data center and external connections to the Internet restricts total server capacity.
Total power input: The electrical input of a data center limits its capacity as only as many servers as can be powered.
Power efficiency: The efficiency at which data center equipment consumes power, typically measured by power usage effectiveness (PUE), is also a significant factor. Higher energy efficiency allows for more workloads to be hosted using a fixed power capacity.
The Pros and Cons of Sizing Data Centers Based on Energy Input
Of all the factors mentioned, power input plays a crucial role in comparing data center capacity. While metrics like square footage and total server rack count are also discussed, the measure of how many megawatts of energy flow into a data center tends to be the go-to indicator of size.
Measuring data center capacity based on power input has its advantages. It provides a relatively consistent proxy for determining how many servers and workloads a data center can support since more workloads require more power, suggesting more capacity.
However, a significant drawback is that total energy input doesn’t always translate linearly to total server capacity. Server power requirements can vary based on components, hosted workloads, and cooling system efficiency. Pairing total power input measurements with PUE metrics can offer a more meaningful assessment as PUE reflects the percentage of energy that can be used to power servers within a data center.
Other metrics like square footage or server rack count are imperfect measurements of a data center’s capabilities. Server and rack sizes can differ, making it challenging to equate square footage to hosting capacity. Even when physical space isn’t a constraint, cooling limitations can pose challenges.
In the absence of a better alternative, the industry may default to focusing on total energy input when assessing data center capacity.
Size Matters: Accurately Calculating Data Center Capacity
In conclusion, accurately measuring data center size involves considering various factors, including energy input, but it’s essential to recognize that assessments of data center capacity only roughly correlate with the number of workloads a facility can host. Data centers are intricate environments, and no single size metric can fully capture their complexity without oversimplifying.
By weighing multiple metrics and understanding the nuances of data center capacity, stakeholders can make more informed decisions when evaluating and comparing data centers.
