Have a Question?

Power Draws for the Vault w/ UPS Uptime Estimates

Print

Power Draws

The following section is intended to help those that are designing a network and need to calculate the power load on a specific circuit or create an energy budget for a new project. The data may also help a Vault owner analyze an overheating/stress issue. 

Test Conditions

All tests were performed on Vaults at an ambient temperature of 74F with either Crucial 8GB DDR3 or DDR4, and either a Protectli 120GB mSATA, Protectli 120GB M.2 or a Kingston 250GB NVMe. No peripherals or USBs were plugged in and the Vaults were set upright on their rubber feet on a workbench with no active airflow over each unit. The Vaults were given a few inches of clearance on each side and nothing was stacked on top of or above them.

Max Power Draw Tests

In the first round of tests, the Vaults used for testing were "cold", or at ambient room temperature. These tests show a higher wattage draw due to there not being any CPU thermal throttling. Without throttling, the cores will function close to their max frequency including a turbo mode if they have one and will draw the most power they are capable of drawing. These test results should be used when budgeting for peak power draws from equipment on a single circuit.

VAULT MODELCPU Average Frequency in MHz @ temp in CCPU Throttling %Ubuntu at Idle (in avg watts)Ubuntu & s-tui (in avg watts)Ubuntu & s-tui & iperf (in avg watts)Peak Draw Ubuntu & s-tui & iperf (in peak watts)Power Supply
FW2B - J30602480 @ambient04.88.48.89.240w Power Supply
FW4B - J31602240 @ambient05.411.011.412.140w Power Supply
VP2410 - J41252600T @ambient07.213.814.515.540w Power Supply
VP2420 - J64122600T @ambient 08.216.018.519.740w Power Supply
VP4630 - 10110U3400T @ambient011.833.434.134.960w Power Supply
VP4650 10210U4200T @ambient010.435.837.144.190w Power Supply
VP4670 - 10810U4900T @ambient09.443.554.257.690w Power Supply
FW6A - 3865U1800 @ambient06.913.213.814.140w Power Supply
FW6B - 7100U2400 @ambient07.016.918.518.960w Power Supply
FW6Br2 - 8130U3400T @ambient07.632.035.136.060w Power Supply
FW6C - 7200UX3100T @ambient07.431.332.934.160w Power Supply
FW6D - 8250U3400T @ambient07.558.460.661.960w Power Supply
FW6E - 8550U3600T @ambient07.870.174.280.190w Power Supply

Heat Saturated Power Draw Test

In the second round of tests the Vaults were measured after running an s-tui stress test. s-tui is a terminal UI that can be used for monitoring your Vault. s-tui allows monitoring of a Vaults CPU temperature, CPU frequency, power and utilization via a graphical user interface (GUI). It was written in Python and doesn't require root privilege to access and use, it uses very little system resources, and it doesn't require configuration files to use its core features.

s-tui can be downloaded on Ubuntu via the console with the command: $ sudo apt install s-tui

To use the stress function within s-tui you will need to download the "stress" package with the command: $ sudo apt install stress.

Once both package are downloaded type "s-tui" in the command line to open the GUI. s-tui starts in Monitor mode. To stress test your CPU click on "Stress" in the upper left hand corner.

s-tui was run on Ubuntu 22.04 "Jammy Jellyfish" for one hour to bring the internal temperatures up to the highest typically seen under a full 100% CPU load known as "heat saturated". The results show a lower power draw on the Vaults that have CPUs with an Intel Turbo mode due to CPU throughput throttling and reflect the amount of throttling the CPU is using to keep the temperature below 78C.

***NOTE: These test conditions are well beyond typical use conditions for a Vault. Most applications are not so CPU intensive that they require constant Turbo.

VAULT MODELCPU Average Frequency in MHz @ temp in CCPU Throttling %Ubuntu at Idle (in avg watts)Ubuntu & s-tui (in avg watts)Ubuntu & s-tui & iperf (in avg watts)Power Supply
FW2B - J3060 w/ 8GB DDR32480 @51C0%4.28.48.640w Power Supply
FW4B - J3160 w/ 8GB DDR32240 @61C0%5.611.011.240w Power Supply
VP2410 - J4125 w/ 8GB DDR42300 @53C11.5%7.313.514.140w Power Supply
VP2420 - J6412 w/ 8GB DDR42600 @72C0%8.116.018.440w Power Supply
VP4630 - 10110U w/ 8GB DDR42600 @64C23.6%12.423.724.360w Power Supply
FW6A - 3865U w/ 8GB DDR41800 @41C0%6.713.313.540w Power Supply
FW6B - 7100U w/ 8GB DDR42400 @55C0%8.117.818.260w Power Supply
FW6Br2 - 8130U w/ 8GB DDR42750 @79C19.2%7.822.422.960w Power Supply
FW6C - 7200UX w/ 8GB DDR42500 @64C20.4%6.919.620.960w Power Supply
FW6D - 8250U w/ 8GB DDR42500 @78C27.5%7.628.830.060w Power Supply
FW6E - 8550U w/ 8GB DDR42470 @78C28.6%8.029.932.190w Power Supply

UPS Uptime Estimates

The following section is intended to help those looking to estimate the uptime their Vault will have upon an AC power loss to their network. Two tests were conducted to test the estimated uptime of a Vault running on our 44.4Wh UPS (Uninterruptible Power Supply) and our 88.8Wh UPS. The first test was performed running the Vaults with OPNsense 22.1 "Observant Owl" at the default configuration with no traffic. This replicates a power loss at home or a small office during off-business hours. The second test was performed running the Vaults with the default configuration OPNsense between two Vaults, one as the iPerf3 server and the other as the iPerf3 client. This allowed us to measure the power draw of a Vault seeing passthrough traffic instead of being the origination or termination of the iPerf stream. This test replicated peak traffic conditions a Vault would experience on a network as an edge-of-network firewall appliance.

The resulting estimated uptime is shown in Watt Hours which is the amount of energy equal to using 1 watt for 1 hour. For example, a Vault drawing 4.4watts will last 10 hours on our 44.4Wh battery: 44.4Wh/4.4watts=10.

OPNsense 22.1 @ Idle Uptime Estimate

This test was performed to estimate the time a Vault will continue running on our 44.4Wh and 88.8Wh UPS once AC power is lost. OPNsense was run with default settings and no additional added packages. No traffic was allowed to pass through to replicate the least amount of power a Vault firewall would draw on an idle network.

OPNsense 22.1 @ Max Throughput Uptime Estimate

This test was performed to estimate the time a Vault will continue running on our 44.4Wh and 88.8Wh UPS once AC power is lost. OPNsense was run with default settings and no additional added packages. Iperf3 traffic was set up from LAN to WAN (running at ~930Mbps) to replicate maximum traffic conditions on a 1Gbps network and the amount of power a Vault would draw.

VAULT MODELApprox Wattage Draw Running OPNsense @ idleEst Uptime 44.4Wh UPS in HoursEst Uptime 88.8Wh UPS in HoursEst CPU Usage Running OPNsense & iPerf3Approx Wattage Draw Running OPNsense & iPerf3Est Uptime 44.4Wh UPS in HoursEst Uptime 88.8Wh UPS in Hours
FW2B - J30607.65.8411.6853%8.85.04510.091
FW4B - J31607.46.0012.0031%9.44.7239.447
VP2410 - J412576.3412.6923%9.14.8799.758
VP2420 - J64129.74.589.1517%11.2 @ 2.5G3.9647.929
VP4630 -10110U12.43.587.1620%16.9 @ 2.5G2.6275.254
VP4650 - 10210U
VP4670 - 10810U
FW6A - 3865U9.64.639.2567%11.63.8287.655
FW6B - 7100U7.95.6211.2432%11.33.9297.858
FW6B - 8130U7.95.6211.2410%10.94.0738.147
FW6C - 7200UX9.34.779.5521%10.94.0738.147
FW6D - 8250U9.54.679.3513%12.03.7007.400
FW6E - 8550U9.74.589.159%12.23.6397.279

Devices used to measure power draws

Kill A Watt EZ P3

Baldr BA3091

Table of Contents