Shop
Support
Community
TechCenter
Home
Topics: All
Wikis
Forums
Blogs
Video
TechChat
Events
About
TechCenter
Dell Community
Search Options
Search Everything
Search Virtualization
TechCenter
>
Virtualization
>
Virtualization - Wiki
>
Advantages of Dell Servers over HP for Virtualization
Join
Sign in
Advantages of Dell Servers over HP for Virtualization
Virtualization
Home
Forum
Wiki
Group and Wiki Navigation
Loading...
Search
Article
History
Advantages of Dell Servers over HP for Virtualization
Virtualization - Wiki
This page is the wiki home for a white paper that compares the price/performance of the Dell™ PowerEdge™ 2950 and HP Proliant DL585 G2 for virtualization. The
complete paper
with all details is linked, and a summary/highlights of the paper are included directly on this page.
Discussion and questions about the paper can be written in the discussion section at the bottom of this page.
Advantages of Dell PowerEdge 2950 2-Socket Servers over HP Proliant DL 585 G2 4-Socket Servers for Virtualization
White Paper Executive Summary
To illustrate the advantages of using two-socket servers for virtualization over four-socket servers, a test was conducted with VMware® Infrastructure 3 on the Dell PowerEdge 2950 and the
HP
ProLiant DL585 G2. The results of these tests show that three PowerEdge 2950 two-socket servers can provide up to 44 percent more performance, 57 percent more performance per watt, and a 95 percent average advantage in price/performance than two HP ProLiant DL585 G2 four-socket servers.
White Paper Highlights
All the tables on this page are taken directly from the white paper. Please see the
complete paper
for full details.
This table from the paper shows the configuration and pricing of the two servers used in the testing:
HP ProLiant DL585 G2
PowerEdge 2950 servers
Virtualization software
VMware ESX Server 3.0.1
VMware ESX Server 3.0.1
Processor
Four dual-core AMD Opteron 2.8 GHz 8220SE processors with 1 MB L2 caches (per core)
Two quad-core Intel Xeon X5355 processors at 2.66 GHz with 8MB cache (shared)
HyperTransport / Frontside bus
1 GHz
1,333 MHz
Memory
32 GB (16x 667 MHz PC2-5300 DDR2 2GB DIMMs)
16 GB (8x 667 MHz fully buffered 2GB DIMMs)
Internal disks
Two Serial Attached SCSI (SAS) 73 GB, 15,000 rpm drives
Two Serial Attached SCSI (SAS) 146 GB, 15,000 rpm drives
Network interface card (NIC)
Two 10/100/1,000 Mbps internal NICs
Two 10/100/1,000 Mbps internal NICs
Disk controller
Smart Array P400
PERC 5/i
Hardware Cost
$28,088
$13,498
VMware Licensing Cost
$13,914
$7,188
Source
http://www.hp.com
12/18/2006
http://
www.dell.com
12/18/2006
This table from the paper shows the performance results of the two servers on three different virtualization workloads:
HP ProLiant DL585 G2
PowerEdge 2950
Workload
Number of VMs
Performance
Power Consumption (W)
Number of VMs
Performance
Power Consumption (W)
Microsoft SQL Server 2005
36
31,729 OPM
719
32
29,346 OPM
449
SUSE LAMP
46
10,093 OPM
743
44
9,852 OPM
447
NetBench
52
1,028 Mb/sec
735
42
1,001 Mb/sec
444
The next table shows how the performance results and associated costs would scale, and as additional Dell two-socket or HP four-socket servers were added to the server farm:
2 HP ProLiant DL585 G2s
3 PowerEdge 2950s
Dell Advantage
Total Number of Sockets
8
6
SQL Server 2005
72 VMs / 63,458 OPM
96 VMs / 88,038 OPM
33% / 39%
SUSE LAMP
92 VMs / 20,186 OPM
132 VMs / 29,556 OPM
44% / 46%
NetBench
104 VMs / 2056 Mb/s
126 VMs / 3003 Mb/s
22% / 46%
Averaged Cumulative Power Consumption on SQL/SUSE/ Netbench (W)
1465
1340
9%
Hardware Cost
$56,176
$40,494
39%
ESX Licensing
$27,828
$21,564
29%
Total Cost
$84,004
$62,058
35%
Performance per watt:
HP ProLiant DL585 G2
PowerEdge 2950
Dell Advantage
SQL Server 2005
44.1 OPM / watt
65.4 OPM / watt
48%
SUSE LAMP
13.6 OPM / watt
22.0 OPM / watt
62%
NetBench
1.4 Mb/s / watt
2.3Mb/s / watt
61%
57 % Average
Price/Performance:
2 HP ProLiant DL585 G2
s
3 PowerEdge 2950s
Dell Advantage
SQL Server 2005
$1.32 / OPM
$ .70 / OPM
88 %
SUSE LAMP
$4.16 / OPM
$2.09 / OPM
98 %
NetBench
$40.86 / Mb/s
$20.67 / Mb/s
98 %
95 % Average
Please post comments and questions about this paper just below in the comments section - Thanks.