The following protocol should be followed when measuring power consumption levels of computers for compliance with the Off, Sleep, and Idle levels provided in this ENERGY STAR Version 5.0 Computer Specification. Partners must measure a representative sample of the configuration as shipped to the customer. However, the Partner does not need to consider power consumption changes that may result from component additions, BIOS and/or software settings made by the computer user after the product is sold. This procedure is intended to be followed in order and the mode being tested is labelled where appropriate.
Computers must be tested with configuration and settings as shipped, unless otherwise specified in the test procedure in this Appendix A. Steps requiring alternative setup are marked with an asterisk (‘*’).
Unless otherwise specified, all terms used in this document are consistent with the definitions contained in the Version 5.0 ENERGY STAR Eligibility Criteria for Computers.
1.UUTUUT is an acronym for ‘unit under test,’ which in this case refers to the computer being tested.
2.UPSUPS is an acronym for ‘Uninterruptible Power Supply,’ which refers to a combination of converters, switches and energy storage means, for example batteries, constituting a power supply for maintaining continuity of load power in case of input power failure.
Approved meters will include the following attributes [Characteristics of approved meters taken from IEC 62301 Ed 1.0: Measurement of Standby Power]:
Power resolution of 1 mW or better;
An available current crest factor of 3 or more at its rated range value; and
Lower bound on the current range of 10 mA or less.
The following attributes in addition to those above are suggested:
Frequency response of at least 3 kHz; and
Calibration with a standard that is traceable to the U.S. National Institute of Standards and Technology (NIST).
It is also desirable for measurement instruments to be able to average power accurately over any user selected time interval (this is usually done with an internal math's calculation dividing accumulated energy by time within the meter, which is the most accurate approach). As an alternative, the measurement instrument would have to be capable of integrating energy over any user selected time interval with an energy resolution of less than or equal to 0.1 mWh and integrating time displayed with a resolution of 1 second or less.
Measurements of power of 0.5 W or greater shall be made with an uncertainty of less than or equal to 2 % at the 95 % confidence level. Measurements of power of less than 0.5 W shall be made with an uncertainty of less than or equal to 0.01 W at the 95 % confidence level. The power measurement instrument shall have a resolution of:
0.01 W or better for power measurements of 10 W or less;
0.1 W or better for power measurements of greater than 10 W up to 100 W; and
1 W or better for power measurements of greater than 100 W.
All power figures should be in watts and rounded to the second decimal place. For loads greater than or equal to 10 W, three significant figures shall be reported.
(Reference IEC 62301: Household Electrical Appliances – Measurement of Standby Power, Sections 4.2, 4.3, 4.4)
Supply Voltage: | North America/Taiwan: | 115 (± 1 %) Volts AC, 60 Hz (± 1 %)
Europe/Australia/New Zealand: | 230 (± 1 %) Volts AC, 50 Hz (± 1 %)
Japan: |
100 (± 1 %) Volts AC, 50 Hz (± 1 %)/60 Hz (± 1 %)
Note: For products rated for > 1.5 kW maximum power, the voltage range is ± 4 %
Total Harmonic Distortion (THD) (Voltage): | < 2 % THD (< 5 % for products which are rated for > 1.5 kW maximum power)
Ambient Temperature: | 23 °C ± 5 °C
Relative Humidity: | 10 – 80 %
Power consumption of a computer shall be measured and tested from an AC source to the UUT.
If the UUT supports Ethernet, it must be connected to an Ethernet network switch capable of the UUT's highest and lowest network speeds. The network connection must be live during all tests.
The AC power consumption of a computer should be measured as follows:
Record the manufacturer and model name of the UUT.
Ensure that the UUT is connected to network resources as detailed below, and that the UUT maintains this live connection for the duration of testing, disregarding brief lapses when transitioning between link speeds.
Desktops, Integrated Desktops, and Notebooks shall be connected to a live Ethernet (IEEE 802.3) network switch as specified in Section II., ‘Test Configuration,’ above. The computer must maintain this live connection to the switch for the duration of testing, disregarding brief lapses when transitioning between link speeds. Computers without Ethernet capability must maintain a live wireless connection to a wireless router or network access point for the duration of testing.
Small-Scale Servers shall be connected to a live Ethernet (IEEE 802.3) network switch as specified in Section II., ‘Test Configuration,’ above, and that the connection is live.
Thin Clients shall be connected to a live server via a live Ethernet (IEEE 802.3) network switch and shall run intended terminal/remote connection software.
Connect an approved meter capable of measuring true power to an AC line voltage source set to the appropriate voltage/frequency combination for the test.
Plug the UUT into the measurement power outlet on the meter. No power strips or UPS units should be connected between the meter and the UUT. For the test to be valid the meter should remain in place until all Off, Sleep, and Idle power data is recorded.
Record the ac voltage and frequency.
Boot the computer and wait until the operating system has fully loaded. If necessary, run the initial operating system setup and allow all preliminary file indexing and other one-time/periodic processes to complete.
Record basic information about the computer's configuration – computer type, operating system name and version, processor type and speed, and total and available physical memory, etc.
Record basic information about the video card or graphics chipset (if applicable) - video card/chipset name, frame buffer width, resolution, amount of onboard memory, and bits per pixel.
* Ensure that the UUT is configured as shipped including all accessories, WOL enabling, and software shipped by default. UUT should also be configured using the following requirements for all tests:
Desktop systems shipped without accessories should be configured with a standard mouse, keyboard and external computer display.
Notebooks should include all accessories shipped with the system, and need not include a separate keyboard or mouse when equipped with an integrated pointing device or digitizer.
Notebooks should have the battery pack(s) removed for all tests. For systems where operation without a battery pack is not a supported configuration, the test may be performed with fully charged battery pack(s) installed, making sure this configuration is reported in the test results.
Small-Scale Servers and Thin Clients shipped without accessories should be configured with a standard mouse, keyboard and external computer display (if server has display output functionality).
For Computers with Ethernet capability, power to wireless radios should be turned off for all tests. This applies to wireless network adapters (e.g. 802.11) or device-to-device wireless protocols. For Computers without Ethernet capability, power to a wireless LAN radio (e.g. IEEE 802.11) should remain on during testing and must maintain a live wireless connection to a wireless router or network access point, which supports the highest and lowest data speeds of the client radio, for the duration of testing.
Primary hard drives may not be power managed (‘spun-down’) during Idle testing unless they contain non-volatile cache integral to the drive (e.g. ‘hybrid’ hard drives). If more than one internal hard drive is installed as shipped, the non-primary, internal hard drive(s) may be tested with hard drive power management enabled as shipped. If these additional drives are not power managed when shipped to customers, they must be tested without such features implemented.
The following guidelines should be followed to configure power settings for computer displays (adjusting no other power management settings):
For computers with external computer displays (most desktops): use the computer display power management settings to prevent the display from powering down to ensure it stays on for the full length of the Idle test as described below.
For computers with integrated computer displays (notebooks and integrated systems): use the power management settings to set the display to power down after one minute.
Shut down the UUT
With the UUT shut down and in Off, set the meter to begin accumulating true power values at an interval of less than or equal to one reading per second. Accumulate power values for five additional minutes and record the average (arithmetic mean) value observed during that five minute period. [Laboratory-grade, full-function meters can integrate values over time and report the average value automatically. Other meters would require the user to capture a series of changing values every 5 seconds for a five minute period and then compute the average manually.]
Switch on the computer and begin recording elapsed time, starting either when the computer is initially switched on, or immediately after completing any log in activity necessary to fully boot the system. Once logged in with the operating system fully loaded and ready, close any open windows so that the standard operational desktop screen or equivalent ready screen is displayed. Between five and 15 minutes after the initial boot or log in, set the meter to begin accumulating true power values at an interval of at least one reading per second. Accumulate power values for five additional minutes and record the average (arithmetic mean) value observed during that five minute period.
After completing the Idle measurements, place the computer in Sleep mode. Reset the meter (if necessary) and begin accumulating true power values at an interval of at least one reading per second. Accumulate power values for five additional minutes and record the average (arithmetic mean) value observed during that five minute period.
If testing both WOL enabled and WOL disabled for Sleep, wake the computer and change the WOL from Sleep setting through the operating system settings or by other means. Place the computer back in Sleep mode and repeat step 14, recording Sleep power necessary for this alternate configuration.
The test results must be reported to the EPA or the European Commission, as appropriate, taking care to ensure that all required information has been included, including modal power values and eligible capability adjustments for Desktops, Integrated Desktops, and Notebooks.
The maximum power for workstations is found by the simultaneous operation of two industry standard benchmarks: Linpack to stress the core system (e.g., processor, memory, etc.) and SPECviewperf® (latest available version for the UUT) to stress the system's GPU. Additional information on these benchmarks, including free downloads, can be found at the URLs found below:
Linpack | http://www.netlib.org/linpack/
SPECviewperf® | http://www.spec.org/benchmarks.html#gpc
This test must be repeated three times on the same UUT, and all three measurements must fall within a ± 2 % tolerance relative to the average of the three measured maximum power values.
Measurement of the maximum AC power consumption of a workstation should be conducted as follows:
Connect an approved meter capable of measuring true power to an AC line voltage source set to the appropriate voltage/frequency combination for the test. The meter should be able to store and output the maximum power measurement reached during the test or be capable of another method of determining maximum power.
Plug the UUT into the measurement power outlet on the meter. No power strips or UPS units should be connected between the meter and the UUT.
Record the AC voltage.
* Boot the computer and, if not already installed, install Linpack and SPECviewperf as indicated on the above Websites.
Set Linpack with all the defaults for the given architecture of the UUT and set the appropriate array size ‘n’ for maximising power draw during the test.
Ensure all guidelines set by the SPEC organisation for running SPECviewperf are being met.
Set the meter to begin accumulating true power values at an interval no more than one reading per second, and begin taking measurements. Run SPECviewperf and as many simultaneous instances of Linpack as needed to fully stress the system.
Accumulate power values until SPECviewperf and all instances have completed running. Record the maximum power value attained during the test.
The test results must be reported to the EPA or the European Commission, as appropriate, taking care to include all required information.
Upon submittal of data, manufacturers must also include the following data:
Value of the n (the array size) used for Linpack,
Number of simultaneous copies of Linpack run during the test,
Version of SPECviewperf run for test,
All compiler optimisations used in compiling Linpack and SPECviewperf, and
A precompiled binary for end users to download and run both SPECviewperf and Linpack. These can be distributed either through a centralised standards body such as SPEC, by the OEM or by a related third party.
This testing procedure describes the method by which a single unit may be tested for compliance. An ongoing testing process is highly recommended to ensure that products from different production runs comply with ENERGY STAR.