series of tests were with the "real world" Business and
Content Creation Winstone benchmarks and a low resolution
time demo run using Quake 3 Arena. These tests don't
demonstrate the true power of a multi-processor system
though, so continue reading until the our conclusion
section, for a simplified explanation why.
Business & Content Creation Winstone
We ran ZD
Labs' Business Winstone 2001 benchmark first. We'll
directly quote ZD's eTestingLabs website for an
explanation as to what this test is comprised of:
"Business Winstone is a system-level, application-based
benchmark that measures a PC's overall performance when
running today's top-selling Windows-based 32-bit
applications on Windows 98 SE, Windows NT 4.0 (SP6 or
later), Windows 2000, Windows Me, or Windows XP. Business
Winstone doesn't mimic what these packages do; it runs
real applications through a series of scripted activities
and uses the time a PC takes to complete those activities
to produce its performance scores."
The Business Winstone tests include:
Microsoft Office 2000 applications (Access, Excel,
FrontPage, PowerPoint, and Word)
The MSI K7D
Master was crushed by the KT333 based single CPU system.
The significant performance delta can be attributed to
three things. One, the lackluster hard drive
performance we saw with AMD's chipset. Two, the
66MHz. clockspeed advantage of the Athlon XP 2100+ and
three, the lower memory bandwidth provided of the AMD 762
Next up is
ZD's Content Creation Winstone 2002. This benchmark
runs a similar series of scripted activities as well, but
the tests are comprised of more "bandwidth hungry"
applications. The applications used in the Content
Creation Winstone 2002 tests include:
Dreamweaver UltraDev 4
Windows Media Encoder 7.01.00.3055
Foundry Sound Forge 5.0c (build 184)
Some of the
applications used in the Content Creation tests are
multithreaded, and this benchmark is not affected as much
by slower hard drive performance, which is why the
difference in scores is much less pronounced.
Nonetheless, the KT333/2100+ combo bests the MSI K7D
Master by about 10% here, but this kind of performance
difference would be unperceivable to the end user.
Quake 3 Arena
No SMP Here...
Quake 3 is one
of the few games that "technically" has the ability to
take advantage of a second CPU. We ran a timedemo at
low-resolution to isolate CPU performance.
no matter what we tried, we were not able to get Quake 3
to run in "SMP Mode" on the MSI K7D Master. We were
relegated to running a "standard" time demo.
Last year I reviewed Abit's Dual-PIII board, the VP6,
was able to get Quake 3 working properly in "SMP mode" and
saw roughly a 20% increase in frame rate. If anyone
has had any luck getting Quake 3 working on their Dual-Athlon
rig, drop us a line
and let us know what you did. I'd like to update
this review with those scores if possible.
SO WHAT DO ALL
THESE BENCHMARKS MEAN, IS A DUAL-CPU SYSTEM ACTUALLY
"SLOWER" THAN A SINGLE-CPU SYSTEM?
answer is absolutely not. After looking at our
benchmark results, many of you may be dismissing a Dual-Athlon
system. You're thinking there is no real benefit to
going with a second CPU unless you're running applications
specifically coded to take advantage of it...but you'd be
There is a
common misconception that a Dual-CPU system will only
outperform a single CPU system when multi-threaded (SMP
compatible) applications are used, but this simply isn't
true. Without giving an overly technical
explanation, think of every application or program running
on your system as being a thread. At any given time
you may have upwards of 29 or 30 single threaded
applications or processes running on your machine (open up
Task Manager and see for yourself, odds are you'll have
even more than that running). On a single CPU system
all of these threads are processed by the lone processor,
while on the Dual-CPU system the threads can be split
between the 2 CPUs. This is a VERY over-simplified
explanation, but it accurately explains the advantage of a
Dual-CPU system. This ability to balance the processing
load across multiple CPUs, in general yields a more
responsive system. Anyone who has spent a
significant amount of time using a Dual-CPU system knows
what we mean.
Winstone 2001 as an example, I'd like you to think about
these benchmark scores another way. Theoretically
speaking, because the applications used in these tests are
not multithreaded, we could have run the Business Winstone
2001 benchmark in two concurrent instances on the Dual-CPU
system, without seeing a huge drop in performance, while a
Single-CPU system's scores would have been halved.
To put things simply, if you multitask, and keep numerous
applications open on your machine at once, you'll
definitely "see" and feel the benefits of a Dual-CPU rig.
MSI K7D Master lacks many of the bells and whistles found
on some "enthusiast" boards, like RAID or 6 Channel audio,
it still managed to impress us on a few different levels.
MSI has targeted this board at the low to mid-range server
and workstation markets, and we feel they have built a
product that caters to them perfectly. Our only
"pseudo-gripe" with the K7D Master is the potential for
some users to have "broken" USB 1.1 ports. This
issue really isn't MSI's fault though. Errata in the
original revision of the AMD 768 Southbridge is the
culprit. Regardless, our board's USB 1.1 ports
seemed to work properly, and MSI includes a 5-Port USB 2.0
PCI card anyway. The overall performance of this
board is great, and stability throughout our battery of
tests was top-notch. The only way we could get this
board to crash was to overclock the board well past 10%
mark, which isn't what the K7D Master was designed to do
in the first place. Based on it's good performance,
rock solid stability and versatility in configuration
options, but because it has limited expandability and a
flakey Southbridge, we give the MSI K7D Master a
HotHardware Heat Meter Rating of...
or any other HotHardware article in the PC Hardware Forum!