By
Jeff Bouton
5/12/2002

 |
Hot Hardware's Test Systems |
Something Old, Something
New... |
|
Test Machine A
MSI KT3 Ultra ARU
Motherboard
AMD AthlonXP
1900+.
512MB Corsair XMS2400
(2-2-5-2)
2 IBM GXP 120 7200RPM
82GB HDs
ATI All-In-Wonder Radeon 8500
D-Link DFE-530TX+ NIC
Creative 52X CD-ROM
Hercules Game Theater
6.1 Sound
Standard Floppy Drive
Windows XP Professional
DirectX 8.1
ATI Display Drivers Ver. 6.13.10.6043 |
Test Machine
B
Shuttle AK35GT2/R
Motherboard
AMD AthlonXP
1800+.
256MB Corsair XMS2400
(2-2-5-2)
2 IBM GXP 60 7200RPM
40GB HDs
eVGA e-GeForce4 Ti4600 w/ACS
Intel Pro/100+ NIC
Creative 52X CD-ROM
On-Board Audio
Standard Floppy Drive
Windows XP Professional
DirectX 8.1
Detonator 28.32 |
Test Machine C
DFI CS60-EC
Motherboard
Pentium III 1GHz.
256MB Mushkin Rev.3
SDRAM
1 Western Digital
WD300B 7200RPM
30GB HD
X-Micro GeForce2 MX
D-Link DFE-530TX+ NIC
Creative 52X CD-ROM
Sound Blaster Live
Standard Floppy Drive
Windows ME
DirectX 8.1
Detonator 28.32 |
 |
The Benchmarks |
Short and Sweet... |
|
Network performance is
not an exacting science, nor are we aware of any
standard when it comes to
benchmarking networking equipment. With various
computers running an inordinate combination of
hardware, it is virtually impossible to come up with
a balanced test routine that would yield consistent
results every time. Nevertheless, we've outlined three test scenarios that transfer a
set amount of data between
three test machines to demonstrate how well the
network performs. Each test was performed
using the Compex DSR2216 DynaStack 16-Port Switch as well as a
NetGear DS524 Hub. To run these tests, we used
a handy command line utility call NetCPS.
NetCPS is unique in the sense that it can transfer a
specific amount of data across a network without
relying on a systems hard drive. Hopefully by
the end of the three tests, we will have a clear
example of the how a switch based network is
superior to a hub based network. Let's get on
with it!
-
Test Scenario 1:
In this test, computer A was configured in to transfer 500MBs
of data to computer B.
-
Test Scenario 2:
Here, both computers A and B were configured to
transfer 500MBs of data simultaneously between each
other.
-
Test Scenario 3:
Finally, computer A was configured to transfer 500MBs of
data to computer B, while computer B was
transferring the same to computer C as well as
computer C to computer A. This put a
fair amount of load on the network and will show the
greatest difference between the Switch and the Hub.
|
In the chart below, we
listed the average Megabytes per second the transfer
ran during each test, followed by the total time it
took to complete the transfer of 500MB of data.
Test
Scenario |
NetCPS Results
Compex
DSR2216 Switch |
NetCPS Results
NetGear DS524 Hub |
1 |
9.91MB/s - 51.45 sec |
10.22MB/s - 48.91 sec |
2 |
9.70MB/s - 51.52 sec |
4.59MB/s - 108.84 sec |
3 |
9.12MB/s - 54.80
sec |
2.83MB/s - 176.68 sec |
Clearly
the virtues of a network running the DynaStack
DSR2216 16-Port switch can be seen above.
With the first test there was very little
difference in performance between the DSR2216 and
the NetGear DS524 Hub.
However, as traffic increased on the network, you
can clearly see that the hub is at a disadvantage
versus the DSR2216. As traffic increased with
each test, more collisions occurred with the NetGear
DS524 hub, severely degrading performance. The Compex
DynaStack DSR2216, on the other hand, maintained a
steady transfer rate through out, barely affected by
the increase in network traffic. Essentially, the Compex
DSR2216 DynaStack Switch dropped roughly 8.5% from
test scenario 1 to test scenario 3, while the NetGear Hub dropped an average of
72% overall!

If you
routinely work on a small to mid-sized network, then
you are already aware of the importance of running a
fast, efficient network. As we've demonstrated
with a few simple tests, the advantages of a switch
based network over a Hub based one is easily seen.
The importance of having good, quality equipment
driving your network cannot be stressed enough.
After reviewing the DynaStack DSR2216 16-Port
Rack-Mountable Switch, we believe that Compex has a
solution for you. The DSR2216 has an excellent
variety of features to insure that the it will have
the ability to easily grow as your networking needs
grow. The DSR2216 is a cost effective 16-Port
switch that retails for around $100 at major
retailers and carries a $10 rebate through the end
of June. When you consider its price point and
features, the DynaStack DSR2216 16-Port
Rack-Mountable Switch offers a lot of value for the
dollar.
For excellent quality, documentation, and overall
performance, we give the DynaStack DSR2216 16-Port
Rack Mountable Switch from Compex a Hot Hardware Heat Meter Rating of an
9.

Are you talkin'
ta me? Get into the HotHardware Conference
Room!
|