Estimate aggregate CPU demand from request rate and measured CPU time per request.
CPU time excludes time waiting on disk, network or locks. The model assumes work distributes perfectly across the entered cores; real scheduling, throttling and parallelism can change capacity. Values above 100% indicate demand exceeding the simplified budget.
01 / INPUT
Your calculation
Enter your values below.
YOUR RESULT · Calculating…
Ready when you are.
Enter your values and select Calculate.
Calculated with unrounded values. Display precision follows each result.
Planning estimate only. Measure your actual workload and validate changes under realistic load before applying a server configuration.
Check the working.
Read the formula and worked example below to understand your result.
CPU time excludes time waiting on disk, network or locks. The model assumes work distributes perfectly across the entered cores; real scheduling, throttling and parallelism can change capacity. Values above 100% indicate demand exceeding the simplified budget.
The formula
CPU demand = requests/s × CPU milliseconds / 1000. Utilization = CPU demand / cores × 100.
A quick example
At 100 requests/s and 10 ms CPU time per request, demand is one core-second each second. Across four cores, the estimate is 25%.
How to use this calculator
Choose the settings that match your scenario, using the units shown beside each field.
Enter requests per second, cpu time per request (ms), available cpu cores.
Calculate, then review the method and assumptions alongside the results.
A few common questions
Can I enter wall-clock response time?+
Not unless all that time is CPU work. Waiting time can make response time much larger than CPU time.
Is this the same as Unix load average?+
No. Load average counts runnable and certain waiting tasks; this tool estimates a CPU-time percentage.