HP ProLiant BL280c HP Power Advisor utility: a tool for estimating power requi - Page 4

Power Advisor. - g6 product number

Get HP ProLiant BL280c - G6 Server PDF manuals and user guides View all HP ProLiant BL280c manuals
Add to My Manuals
Save this manual to your list of manuals

Page 4 highlights

the table of Figure 1, the total amperage and VA figures based on the nameplate ratings of the 750W power supply suggest that substantial power distribution components would be necessary, particularly if power redundancy is required. Figure 1. Rack loading with 20 ProLiant DL380 G6 servers using 750W power supplies (PSs) Wattage per PS unit AC input current per PS unit (@ 208VAC) Total rack wattage Total rack current Total input VA (estimated) Nameplate ratings-based 750 W 4.5 A 15kW 90 A 15 - 17kVA Actual operating needs 300 W 1.38 A 6kW 27.6 A 6 - 7kVA However, measurements of test configurations reveal that the actual operating needs of a given DL380 G6 configuration can be less than half the power that nameplate ratings of the 750W power supply indicate. This means fewer power distribution components are needed, and it reveals the possibility of a better power supply solution. The efficiency of a power supply is determined by how much AC input power is needed to produce a given amount of output power. A power supply requiring 300 watts input to produce 250 watts of output is operating at approximately 83 percent efficiency (250 / 300 = .83). The 50-watt energy delta between the input and output is lost as heat, which must be removed by the cooling equipment. Power supply efficiency is not linear or flat across the output range, and most power supplies achieve maximum efficiency when they operate in the mid to upper range of their rated capacity. A 750-watt power supply providing 300 watts (40 percent capacity) is less efficient than a 460-watt power supply providing the same 300 watts of power (65 percent capacity). In choosing the most efficient (right-sized) power supply for a server, an accurate power consumption estimate for that server is important, particularly for a large (enterprise) data center where cumulative energy losses from a number of servers can be significant. The most accurate power consumption predictions are those obtained by pre-configuring and measuring actual systems under load. This method of obtaining data is usually impractical for customers since it would require purchasing, setting up, configuring, and running of each component to acquire the measurements. HP has tested server products under various configurations and loads to determine actual power requirements. The results of these tests are the foundation in creating the HP Power Advisor. 4

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
the table of Figure 1, the total amperage and VA figures based on the nameplate ratings of the
750W power supply suggest that substantial power distribution components would be necessary,
particularly if power redundancy is required.
Figure 1.
Rack loading with 20 ProLiant DL380 G6 servers using 750W power supplies (PSs)
Nameplate
ratings-based
Actual
operating needs
Wattage per PS unit
750 W
300 W
AC input current per PS unit
(@ 208VAC)
4.5 A
1.38 A
Total rack wattage
15kW
6kW
Total rack current
90 A
27.6 A
Total input VA (estimated)
15 – 17kVA
6 – 7kVA
However, measurements of test configurations reveal that the actual operating needs of a given
DL380 G6 configuration can be less than half the power that nameplate ratings of the 750W power
supply indicate. This means fewer power distribution components are needed, and it reveals the
possibility of a better power supply solution.
The efficiency of a power supply is determined by how much AC input power is needed to produce a
given amount of output power. A power supply requiring 300 watts input to produce 250 watts of
output is operating at approximately 83 percent efficiency (250 / 300 = .83). The 50-watt energy
delta between the input and output is lost as heat, which must be removed by the cooling equipment.
Power supply efficiency is not linear or flat across the output range, and most power supplies achieve
maximum efficiency when they operate in the mid to upper range of their rated capacity. A 750-watt
power supply providing 300 watts (40 percent capacity) is less efficient than a 460-watt power
supply providing the same 300 watts of power (65 percent capacity). In choosing the most efficient
(right-sized) power supply for a server, an accurate power consumption estimate for that server is
important, particularly for a large (enterprise) data center where cumulative energy losses from a
number of servers can be significant.
The most accurate power consumption predictions are those obtained by pre-configuring and
measuring actual systems under load. This method of obtaining data is usually impractical for
customers since it would require purchasing, setting up, configuring, and running of each component
to acquire the measurements. HP has tested server products under various configurations and loads to
determine actual power requirements. The results of these tests are the foundation in creating the HP
Power Advisor.
4