Provisioning ceilings and a source unit discrepancy
Read the explanation
The saved server planner stores four hundred twenty-eight thousand six hundred sixty-six concurrent players and one point two five headroom. Multiplying gives five hundred thirty-five thousand eight hundred thirty-two point five buffered players. Dividing by three hundred twenty and rounding upward requires sixteen hundred seventy-five instances. Bars compare raw and buffered players at point zero zero one pixels each. These are scenario inputs, not actual Steam telemetry or an observed deployment. Instance partition and costs are fixed assumptions. The function multiplies player count by stored sixty-four kilobits per second but divides by one hundred thousand and labels the result gigabits per second, about two hundred seventy-four point three five. Standard decimal conversion from kilobits to gigabits would divide by one million, giving about twenty-seven point four three. Bars compare both labeled quantities at two pixels per gigabit. This source unit discrepancy is preserved, not silently corrected. Neither amount is measured network traffic; no server or provider is provisioned. At sixty-four tick rate the source prices each instance at point seventy-two dollars per hour, giving twelve hundred six compute cost. Hypothetically doubling tick rate to one hundred twenty-eight doubles compute cost to twenty-four hundred twelve while required instances stay unchanged. Bars compare at point two pixels per hourly dollar. Separate egress cost uses the source bandwidth value and a fixed coefficient. Native complete stable JSON report and reset preserve original source; hardware capacity, throughput, costs and advertised game facts remain unverified.