Skip to content
ServerCalc ServerCalc
Power & Energy

HP Server Power Consumption Calculator

Power draw, heat and running cost for HPE ProLiant, calibrated against published SPECpower results.

Inputs

Server

Loads that model's chassis, socket, memory and PSU defaults.

Totals scale across the rack or fleet.

2U chassis · 2 socket(s) · up to 24 drive bays

Processors & memory

TDP drives most of the dynamic power.

W

Empty slots draw nothing.

512 GB installed, about 50 W

Storage & accelerators
W

NICs, HBAs, DPUs and RAID controllers not already counted.

Power supply & load
W

Rating of the supplies actually carrying load: one unit in a 1+1 hot-spare pair, the whole active bank on a GPU node. Nameplate, not draw.

Sets utilization. Pick Custom to drag the slider.

Three-phase carries the same watts at 1/√3 of the line current.

Real power over apparent power. Sets the watts-to-VA gap.

40%
W

Read it off iDRAC, iLO, XCC, IPMI or a metered PDU. Overrides the estimate and recalibrates the whole curve. Leave at 0 to estimate.

Operating profile & cost

24 for always-on production.

$/kWh

Your blended rate per kWh, taxes included.

Facility overhead multiplier. 1.0 counts IT load only.

Grams of CO2e per kWh on your grid.

Rack & facility

Defaults to a full 42U rack minus 2U for switching.

The breaker feeding the rack PDU.

sq ft

Optional. Gives watts per square foot for the room.

kg

kgCO2e to build one server, from the vendor's Product Carbon Footprint report.

Years the manufacturing carbon is amortized over.

HPE ProLiant defaults follow the published QuickSpecs and SPECpower figures for these platforms with the Balanced Power and Performance ROM setting.

  • A full rack peaks at 20761 W, more than the 4892 W a single circuit delivers. Budget 5 circuits per rack.
Results update live as you type.

Results

Estimated power draw
542 W
ProLiant DL380 Gen11 at 40% CPU utilization
Idle draw
266 W
Powered on, no workload.
Max draw (100% load)
1.04 kW
Sustained worst case, not a boost spike.
PSU efficiency here
95.8%
52% of the 1000 W nameplate
Annual energy (IT)
4,747 kWh
24 h/day × 365 days/year
Energy cost per month
$59.34
Energy cost per year
$712
Facility power (PUE 1.50)
813 W
Facility cost per year
$1,068
IT load plus cooling and distribution overhead.
Heat output
1,849 BTU/hr
Cooling required
0.15 tons
One ton of cooling is 12,000 BTU/hr.
Current draw
2.7 A
At 208 V, power factor 0.98.
Apparent power
553 VA
Minimum UPS size
691 VA
Apparent power plus 25% headroom.
Operational carbon
2,734 kg CO2e/yr
384 g CO2e/kWh applied to facility energy.
Manufacturing carbon
263 kg CO2e/yr
1,313 kg per server over 5 years.
Total carbon
2,997 kg CO2e/yr
Operational plus manufacturing. 9% of it is manufacturing.
Servers per circuit
4
30 A at 208 V Single-phase, 80% derate, 4,892 W usable.

Per rack

Rack draw at this load
10.8 kW
20 servers per rack.
Rack draw at 100%
20.8 kW
The figure the PDU and breaker have to survive.
Circuits per rack
5
Rack space used
40U
Of 42U available.

Where the watts go

542 W from the wall

At 40% CPU utilization, per server.

CPUs
347 W · 64%
Memory
50 W · 9%
Storage
37 W · 7%
Board, BMC & I/O
56 W · 10%
Fans
30 W · 5%
PSU conversion loss
23 W · 4%

Power vs CPU utilization

Idle is 26% of max draw

Server power is not proportional to work done. A box at 0% utilization still burns its idle floor, which is why consolidation beats tuning: switching a server off saves the whole curve, tuning it saves only the sloped part.

03136259381.3k0%25%50%75%100%Your setting: 542 WCPU utilizationWall power
Wall powerIdle draw: 266 WMax draw (100% load): 1.04 kW

About this calculator

HP's server line has been HPE ProLiant since the 2015 split, so if you are searching for HP server power figures, ProLiant is what you want. This calculator covers the models still in the field in volume: the 1U DL360 Gen10 and Gen11, the 2U DL380 Gen10 and Gen11, the single-socket EPYC DL325 and DL345 Gen11, the ML350 Gen11 tower, and the entry DL20 Gen11.

HPE's own Power Advisor is a good tool and worth using for a purchase you are about to sign off. It is also a download, it is Windows-first, and it only knows HPE. This does the same arithmetic in a browser, shows the formula, and lets you compare a ProLiant against a PowerEdge or a Supermicro without changing tools.

The ProLiant line is also where this calculator's accuracy is pinned. A DL380 Gen11 with two Xeon Platinum 8592+, 256 GB of DDR5 and a 1000 W Titanium supply comes out at 228 W idle here. HPE's published SPECpower_ssj2008 result for that exact configuration measured 224 W. Every other model in the catalogue rests on the coefficients that anchor produced.

Better still, read the real figure off the machine. iLO reports it directly under Power & Thermal, including present, average and peak draw. Put that in the Measured draw field and the whole curve recalibrates to your server.

The formula

Wall power = (CPUs + memory + drives + GPUs + board + fans) ÷ PSU efficiency at that load

Each component contributes an idle floor plus a share that scales with CPU utilization, and the PSU efficiency comes off the 80 PLUS curve at the actual load ratio rather than the headline percentage.

ProLiant-specific things that move the number:

  • ROM Power Profile. The Balanced Power and Performance default keeps C-states available. Switching to Maximum Performance disables them along with core parking, and adds meaningful idle draw. Static High Performance is worse again. The defaults here assume the balanced profile.
  • Power Regulator and the separately configurable Dynamic Power Savings mode change how aggressively the CPU drops frequency, which shows up directly in the mid-range of the utilization curve.
  • Flexible Slot (Flex Slot) supplies come in 500 W, 800 W, 1000 W, 1600 W and 1800 W, Platinum or Titanium. Fitting a 1600 W unit to a server that draws 300 W parks it below 20% load where even Titanium efficiency falls off. Pick the smallest supply that covers your real maximum with headroom.
  • HPE Power Regulator with dynamic capping can enforce a ceiling. Nothing here models a cap, so a capped server will measure lower than the maximum shown.

Facility figures follow: BTU/hr = watts × 3.412, amps = watts ÷ (volts × PF) single-phase or ÷ (√3 × volts × PF) three-phase, kWh/year = watts × hours × days ÷ 1000.

Common use cases

  • Sizing PDUs, breakers and UPS for a ProLiant rack before the purchase order
  • Choosing between a 500 W, 800 W and 1600 W Flex Slot supply on efficiency at real load
  • Comparing a DL380 Gen10 refresh against a Gen11 on annual electricity cost
  • Getting the BTU/hr figure a facilities team needs for a room of ProLiant nodes
  • Working out how many DL360 nodes fit on one circuit at the 80% continuous-load derate
  • Turning a single iLO reading into annual cost, heat load and carbon

Frequently Asked Questions

Is it HP or HPE?
HPE. Hewlett-Packard split in November 2015: HP Inc. took PCs and printers, Hewlett Packard Enterprise took servers, storage and networking. Every ProLiant server sold since then is an HPE product, though plenty of people still say and search for HP servers, and older Gen8 and Gen9 ProLiants genuinely carry HP branding. This calculator covers the ProLiant line under either name.
How do I check the actual power draw of an HPE ProLiant?
iLO reports it directly. In iLO 5 or iLO 6, go to Power & Thermal, then Power Meter: it shows present reading, plus average, maximum and minimum over 24 hours and 20 minutes, taken from the power supply telemetry. Over the command line you can use `ipmitool sensor` or query the Redfish endpoint /redfish/v1/Chassis/1/Power. That reading is more reliable than any estimate, including this one; put it in the Measured draw field and everything downstream recalibrates.
How much power does an HPE DL380 Gen11 use?
It depends heavily on the CPUs. With two Xeon Platinum 8592+ (350 W each), 256 GB of DDR5 and a Titanium supply, its published SPECpower result measured 224 W idle and 770 W at full benchmark load. With mid-range Xeon Gold 6430s and a typical 16-DIMM production build it idles nearer 260 W and lands around 500 W in normal service. The 1U DL360 Gen11 runs roughly 30 to 50 W below the equivalent DL380 because it has fewer drive bays and a smaller fan wall.
Which Flex Slot power supply should I pick?
The smallest one that covers your maximum draw with about 20 to 30% headroom. Efficiency on the 80 PLUS curve peaks near 50% load and falls off sharply below 20%, so an oversized supply costs real money on a lightly loaded server: a 1600 W Titanium unit carrying 300 W runs near 90% efficient, while an 800 W unit at the same load runs about 95%. Set the PSU capacity field to each option and compare the PSU efficiency here row. Only go large if you plan to add GPUs later.
Do HPE ProLiant servers with two power supplies draw twice the power?
No. A redundant pair carries the same total load, and total draw is essentially unchanged apart from a few watts of standby on the second unit. What redundancy changes is efficiency: when both supplies share a load, each runs at half the load ratio, pushing both into a less efficient part of the curve. HPE's high-efficiency mode addresses this by keeping one supply in standby. The efficiency model here assumes that behaviour, matching a single active supply.
How do I calculate BTU and cooling for a ProLiant rack?
Multiply real wall draw by 3.412 for BTU/hr, then divide by 12,000 for tons of refrigeration. Use actual or realistic maximum draw, never the PSU nameplate, and never the maximum heat dissipation figure from QuickSpecs, which is derived from maximum input power at full load and overstates a real build substantially. A rack of ten DL380 Gen11 nodes at 500 W each is 5 kW, about 17,000 BTU/hr, or 1.4 tons of cooling.
Does this cover ProLiant Gen9, Synergy or Apollo?
Gen10 and Gen11 rack models are in the dropdown. Gen9 is not, but you can approximate one closely by selecting the DL380 Gen10 and setting the CPU to a Xeon E5-2600 v4 part, which is in the CPU list. Synergy and BladeSystem chassis are not modelled, because on those the chassis-level power and cooling is shared across blades and does not decompose cleanly per node. Apollo high-density and the Cray/HPC line are not included. Ask via the feedback form if you need one added.

Spot an error? Have feedback?

Tell us what is wrong with the math, what is missing, or which server model you would like added. We read everything.