Well, in order to test the MxPro I had to take out my 233 MHz processor.
When running the MxPro, my system appeared to be running just as fast as
it was with my 233 MHz chip. My system booted up at 200 MHz and all my
applications appeared to run just as fast.
I then ran Wintune and what I found was interesting. Wintune was
identifying the MxPro as a 486 running at 169 MHz. I found that by
setting my systems clock multiplier to 3.0X, the Wintune results then read
170. "What is going on?" I asked myself. Apparently the
WinChip found in the MxPro is so new that the various benchmark programs
do not know how to identify and test it properly. This interesting
fact can be seen in the chart below.
| Benchmark Program |
MxPro 180 |
MxPro 200 |
| Wintach 1.2 |
114.53 |
131.56 |
| Norton SI 7.0 |
137 |
152 |
| Wincheckit 4.0 |
153 |
170 |
| Wintune 97 |
153 |
170 |
| Norton SI 32 (Win 95) |
157 |
174 |
| Speed 600 |
187 |
210 |
So how do we determine if the MxPro is increasing the performance of my
system? Well, first of all, with this upgrade, many of you will be
upgrading from 75 or 100 MHz. The performance increase will be noticeable
to you the first time you run your system.
Also, Evergreen included a utility to benchmark your system before and
after the upgrade. For example, my boss's PB 600 motherboard, 100
MHz with no L2 cache, went from 67,303 Dhrystones to 127,268 Dhrystones
with the 200 MHz Evergreen MxPro (TM).
To really tell how well the MxPro performed, I had to load Winbench
on my system. This is an extensive benchmark utility that tests
the computer in real world examples of the most popular software titles
available today. The IDT Winchip in my Evergreen scored quite well,
though unfortunately I do not have anything to compare it to other than
my 233 MHz processor. The MxPro I tried scored just behind the 233
MHz Processor, as I expected.
 |
This Winbench test was done on the
same standard PB 640 system, with 256k L2 cache, 48 MB RAM and a Stealth
II S220 video card. Performance will vary between systems.
Keep in mind that a Pentium 100 MHz will score somewhere around 200
(test results included with Winbench 96 are from a 100 MHz Dell Dimension),
and a 233 MHz Packard Bell Multimedia S606 will score around 285 for the
32-bit tests, with no L2 cache. These Benchmark
results can be seen on this
web page at ZD Net.
From the testing and scores I could come up with, it shows the MxPro
is competitive with the Intel chips. A 200 MHz MxPro on a system
with L2 cache can outperform a 233 MHz Intel processor with no L2 cache.
The MxPro should also score well on systems with no L2 cache, since the
MxPro has more L2 cache than an Intel Pentium chip has.
Unfortunately my data and testing hardware are limited to what I have
on hand. I would, however, say the Evergreen MxPro
(TM) performed quite well in all my testing, and the MxPro
was just as stable and almost as fast as my Pentium 233 MMx processor.
The biggest advantage, however, is the MxPro is about $100 less than my
233 MHz processor and performed almost as well.