ISO IEC 23544 pdf – Information Technology — Data centres — Application Platform Energy Effectiveness (APEE)
EXAMPLE 1 A benchmark that executes transactions while increasing the multiplicity untilthe system cannot keep the specified response time in order to obtain maximum multiplicity is notapplicable because the processes of measurement target for these benchmarks have the potential tochange from run to run.
5) Start date and time, and end date and time of all the measurement period(s) are clear.
NOTE2 Some benchmarks have multiple measurement periods in one benchmark execution.
EXAMPLE2(Example of clarification]) A benchmark has warmup and cooldown phases which do noteffort a benchmark performance score. It records start date and time and end date and time of thesephases in a log file, so these phases are distinguishable from measurement period[s].
6) Measurement period start and stop time are measured in seconds or higher precision.7) To ensure that the result is verifiable, the benchmark is made publicly available.
b) The measurer determines the target IT equipment,0Ss and middleware that are to serve as
the elements of the application platform. The application platform shall meet the followingrequirements:
1) The application platform contains all of the target IT equipment necessary to execute the
benchmark.
2) Every target IT equipment component can be measured for energy consumption.6.2 Determination of application outcome
For the determination of application outcome, the following terms are introduced:- One application outcome is defined as one valid execution of the benchmark.The following requirements for measurement shall be met:
a) The benchmark is executed at least one time. If the benchmark is executed multiple times, the
number of benchmark executions is a positive integer.
NOTE The tester can interact with the server between benchmark iterations, typically to calculate thetotal measured time for benchmarks which include unmeasured periods.
b) If the total time of the whole measurement period is less than one hour,the benchmark is repeatedly
executed until the total time of the whole measurement period exceeds one hour.
c) lf the benchmark has a setting for official or valid or reportable results for publication or use in
comparisons,this mode shall be used for APEE comparisons.
6.3Determination of energy consumption of the application platform
6.3.1 Requirements for obtaining energy consumption of the application platform
Energy consumption of the application platform shall be obtained by the sum of energy consumption(kWh) of all the target IT equipment in the application platform during all benchmark measurementperiod[s) of the APEE measurement. Non-benchmark software should be minimal during themeasurement period(s). Execution of non-benchmark software on the application platform shall notinvalidate energy results. Energy measurements during benchmark execution shall not be adjusted toremove the effects of non-benchmark software.
NoTE The application platform is usually used solely by the benchmark during the measuring period foraccuracy.
The energy consumption of each item of target IT equipment in the application platform is obtained bythe measurement method. Details of this method are given in 6.3.2.
6.3.2 Requirements for the measurement method for obtaining energy consumption
The measurement method shall meet the following requirements. The measurer shall replace “SEEM”or “SERTv2”with”APEE” and “SUT”with “application platform”in the text of the SEEM referencedbelow and apply.
a) Environment
1) ISO/IEC 21836:2020,7.2 applies for measurement environment.2) ISO/IEC 21836:2020,7.4 applies for power requirements.
b) Power measurement
1) One or more power analysers shall be used to measure the entire energy consumption of the
application platform.
2) Maximum sampling interval of the power analysers shall be at least one sample per second.3) ISO/IEC 21836:2020,7.3 and 9.2.2.2 apply for power analyser requirements.
4) Power analysers shall comply with ISO/IEC 21836:2020,9.2.2.3 and 9.2.2.4.
5) The following requirements shall be preferentially followed over the above ISO/IEC 21836
requirements.
i)The energy consumption shall be measured by using one or more power analysers.
ii) All clocks of all power analysers and target IT equipment shall be synchronized such that
time stamps are aligned within 1 second.
NOTE This requirement is intended to support measurements of the energy consumption ofmultiple target IT equipment. ISO/IEC 21836 targets a server, but this document targets multiple targetIT equipment.
c)Temperature measurement
1) The instantaneous temperature of an inlet of one typical element of the application platform
shall be measured by a temperature sensor both within 1 minute of the start time of thebenchmark execution and within 1 minute of the end time of the benchmark execution.
2) ISO/IEC 21836:2020,9.2.3 applies for temperature sensor requirements.
d) Approved measuring instrument
1) ISO/IEC21836:2020,Annex B,Table 4 and Table 5 apply for pre-approved power analysers and
temperature sensors.
www.standard-codes.com