CDCP Capacity Planning and Power Budgeting — Questions and Answers
Question 1: What is the formula for Power Usage Effectiveness (PUE), the primary metric for data center energy efficiency?
- PUE = IT Equipment Power / Total Facility Power
- PUE = Total Facility Power / IT Equipment Power (Correct answer)
- PUE = Cooling Power / IT Equipment Power
- PUE = IT Equipment Power / (Cooling Power + Lighting Power)
Correct answer: PUE = Total Facility Power / IT Equipment Power
PUE = Total Facility Power ÷ IT Equipment Power. A perfect PUE of 1.0 would mean all power goes directly to IT equipment with zero overhead. Typical modern data centers target below 1.5; hyperscalers often achieve 1.1–1.2. Values above 2.0 indicate significant inefficiency in cooling and power distribution.
Question 2: In data center capacity planning, what is 'stranded power'?
- Power that is lost due to UPS inefficiency during battery discharge
- Allocated power capacity that cannot be used because cooling or space constraints are reached first (Correct answer)
- Emergency generator capacity reserved exclusively for fire suppression systems
- The difference between utility feed capacity and installed UPS capacity
Correct answer: Allocated power capacity that cannot be used because cooling or space constraints are reached first
Stranded power occurs when power capacity has been allocated (reserved) but cannot actually be deployed because another constraint — typically cooling capacity or physical floor space — has been reached first. It represents wasted capital investment and is a key problem that DCIM tools and capacity planning aim to prevent.
Question 3: A data center operator wants to increase average rack density from 5 kW to 10 kW per rack. Which infrastructure element most commonly becomes the first bottleneck?
- Structured cabling — higher density requires more fiber runs per rack
- Cooling — removing heat from high-density racks exceeds conventional CRAC unit capacity (Correct answer)
- Physical security — higher value equipment requires additional access control layers
- Fire suppression — higher density triggers more frequent false alarms
Correct answer: Cooling — removing heat from high-density racks exceeds conventional CRAC unit capacity
Cooling is the most common bottleneck when rack density increases. Traditional CRAC/CRAH units, raised-floor airflow, and hot/cold aisle containment are designed for lower densities. High-density racks (above ~10 kW) typically require targeted solutions such as in-row cooling, rear-door heat exchangers, or liquid cooling to remove heat effectively.
Question 4: What does 'white space' refer to in data center capacity planning terminology?
- Unpainted concrete areas used for future raised-floor expansion
- The portion of the computer room floor available for IT equipment deployment (Correct answer)
- Empty rack units (U-spaces) within already-installed cabinets
- Cooling headroom measured in kilowatts above current IT load
Correct answer: The portion of the computer room floor available for IT equipment deployment
White space is the customer-facing, revenue-generating floor area of a data center used for IT equipment — as opposed to 'gray space' (mechanical and electrical support areas) or 'white space on paper' (future buildout). Capacity planners track white space utilization to forecast when additional floor space will be needed.
Question 5: A data center is planning to add 200 new servers consuming an average of 400W each. Before approving the deployment, what capacity check is mandatory?
- Verifying that the building permit allows the additional weight load
- Confirming that available power, cooling, space, and network port capacity can support the addition (Correct answer)
- Ensuring the fire suppression system is re-certified for the new equipment
- Checking that the structured cabling standard is upgraded to Cat 6A
Correct answer: Confirming that available power, cooling, space, and network port capacity can support the addition
A proper capacity check before any significant deployment must validate all four constraints: power (80 kW total new load must fit within available PDU and UPS headroom), cooling (thermal load increase must be absorbable), physical space (rack space and floor loading), and network (switch ports, patch panels). Approving only one dimension without checking the others leads to stranded capacity or deployment failures.
Question 6: A data center has a total facility power draw of 2.4 MW and an IT equipment load of 1.6 MW. What is its PUE, and how would it be classified?
- PUE = 0.67 — excellent efficiency
- PUE = 1.5 — average efficiency for a commercial data center (Correct answer)
- PUE = 2.4 — poor efficiency requiring immediate remediation
- PUE = 1.5 — poor efficiency; world-class facilities achieve below 1.2
Correct answer: PUE = 1.5 — average efficiency for a commercial data center
PUE = 2.4 MW / 1.6 MW = 1.5. A PUE of 1.5 is considered average for a commercial data center — not excellent, but not critically poor. World-class hyperscale facilities achieve 1.1–1.2. A PUE of 1.5 means that for every 1W delivered to IT equipment, an additional 0.5W is consumed by power distribution, cooling, and other overhead.
CDCP Capacity Planning and Power Budgeting
What is the formula for Power Usage Effectiveness (PUE), the primary metric for data center energy efficiency?