Dell POWEREDGE M915 - Manual (Page 6)

Power Efficiency Comparison of the Dell PowerEdge M915 and HP ProLiant BL680c G7 Blade Solutions
6
Executive Summary
Introduction
Dell commissioned its Solutions Performance Analysis team to compare the power efficiency of full-chassis
blade solutions using 4-socket blades from Dell and Hewlett-Packard (HP).
In July 2011, the Dell PowerEdge M915 blade server achieved the highest SPECpower_ssj2008 score of any 4-
socket server on the market
1 , so eight of these blades and a PowerEdge M1000e enclosure were used as the
PowerEdge solution.
For the ProLiant solution, HP’s similar BL685c G7 could have been chosen for this study. However, the BL680c
G7 is HP’s top-selling 4-socket blade
2 , so it was chosen to allow examination of HP’s claims of that server’s
unique 4P blade performance and improved power efficiency 3 . Due to its full-height, double-wide form factor,
only four of these blades can fit in a 10U BladeSystem c7000 blade enclosure.
Using the industry-standard SPECpower_ssj2008 benchmark, the performance, power draw, and
performance/watt of full-chassis configurations of both blade solutions were tested. The blade servers were
compared configured as similarly as possible given their architectural differences, with identical memory and
hard drive selections, and best known CPU choices for highest performance/watt ratio. To ensure the lowest
power draw, both blade chassis were equipped with six of their Platinum-rated power supplies for all tests.
The results showed the Dell solution using 8 PowerEdge M915 costs less than the HP solution consisting of just
4 BL680c G7 blades (the maximum that can fit in the BladeSystem c7000 enclosure). The Dell solution also
consumed less power at idle, provided higher raw performance, and achieved a better performance / watt
ratio. The Dell solution also fits twice as many servers in a 10U blade enclosure.
Key Findings
Performance/Watt
 The Dell solution with 8 PowerEdge M915 blades achieved a 92% higher performance / watt ratio
across all load levels than the HP solution using 4 ProLiant BL680c G7 blades.
Performance
 The chassis full of 8 PowerEdge M915 blades achieved 87% higher raw performance at 100% utilization
than the chassis full of 4 HP ProLiant BL680c G7 blades.
Power
 Despite having 87% greater performance, the Dell solution with 8 PowerEdge M915 blades consumes
only 14% more power at 100% utilization compared to the HP solution with just 4 ProLiant BL680c G7
blades.
 Despite having twice as many servers, the Dell blade solution consumes 35% less power at idle than
the HP blade solution.
1
In the published result, the 8-node M915 achieved 14,793,524 ssj_ops@100% target load and a SPECpower_ssj2008 result of 2,716 overall ssj_ops/Watt.
Benchmark results based on results published at www.spec.org as of July 2011. For the latest SPECpower_ssj2008 benchmark results, visit
http://www.spec.org/power_ssj2008/results/power_ssj2008.html.
2 According to IDC Q1 2011 Server Tracker, May 22, 2011
3
http://h10010.www1.hp.com/wwpc/us/en/sm/WF05a/3709945-3709945-3328410-241641-3722793-4268690.html