



                   ͻ
                    ZyXEL Outperforms Other Modems 
                   ͼ



                        ZyXEL Outperforms

In a recent head to head modem evaluation conducted by Daniel
Briere - "...president of TeleChoice, Inc., a Verona, N.J.,
consultancy specializing in strategic planning and analysis of
intelligent networks,services and applications", tested along with
six other leading brand names, "ZyXEL's U-1496Eplus fared the best
across the broad spectrum of lines." (Briere, p. 40, NETWORK WORLD,
MAY 10, 1993)  The lab conducted a series of synchronous and
asynchronous tests on these modems that are the fastest on the
market and found significant differences in the performance of all
the modems across the range of test conditions.

During the test the lab evaluated the performance of modems that
operate at speeds higher than the V.32bis 14.4Kbps. Each modem was
subjected to three tests: a synchronous bit error rate test (BERT),
an asynchronous throughput test transferring the same file pattern
over various network line environments and an asynchronous
throughput test of five common file formats over a clear
transmission line.  Where applicable, testing was performed in
duplex mode.

To assure consistency across the testing procedures, the
performance of the public network was simulated with specialized
equipment from Telecom Analysis Systems Inc. (TAS) A pair of each
modem was connected to a TAS Model 240 Voice band Subscriber loop
emulator which implements test conditions for V.32, V.32bis and
V.Fast modems as outlined by the Electronic Industries
Association/Telecommunications Industry Association (EIA/TIA)
TR30.2 Committee.

ZyXEL Handles Most line Conditions:
ZyXEL's U-1496Eplus passed a vast majority of the synchronous tests
and proved it can handle most line conditions in the United States.
Motorola Codex's 3261Fast with its V.Fast technologies glided
flawlessly across all 59 lines without failing a test (but with
V.Fast a better performance is assured) However, at 19.2 ZyXEL
passed 54 lines of the 59 lines tested thus proving to be the best
in 19.2 modem category. 

To asses bit error rates at the maximum line rates offered by each
of the modems a BERT that sent 10,000 blocks, each 1Kbit in size
over each of the 59 lines was used.  The BERT's were designed to
determine the ability of the modems to effectively convert the
digital information at the modem/data terminal interface to analog
information at the modem/telephone interface and vice versa.  Each
modem manufacturer claimed a particular maximum line rate possible
with their modem and the line speed of each modem was not allowed
to fall back from the maximum.  The modem's terminal flow control
was disabled and a 2047 file pattern was used for the transfer
itself.


"The ZyXEL Modem is Faster...Practically Across The Board"
The asynchronous test across a wide variety of line conditions was
the most important and practical test since it represents the
reliability of a modem through severe line conditions.  In other
words it shows in reality what you could expect from a modem.

Throughput of 10,000 blocks containing 100 characters each was
tested.  In order not to obscure the results in any way the data
compression capabilities were turned off and the 2047 pattern was
used again .  The Terminal flow was set to Request to Send/Clear to
Send(RTS/CTS).

Averaging out as the best in the 19.2 category, ZyXEL's throughput
averaged at 22.7K through all the lines with the lowest deviation
between the best and the poorest, and 21.5K-23K on average of each
type of lines.  Test Result 3 shows that ZyXEL outperformed Codex's
3261 V.Fast modem in 4 line categories.

"Motorola Codex's 3261Fast achieved the highest throughput for most
lines, averaging around 26Kbit/sec." (Briere p.42) But the 3261
could not handle the conditions found in the Logical Loop 1.  "Loop
1 is found in approximately 34% of the nation's Local Loops."
Codex's 3261 averaged 16.4Kbit/sec in Loop 1 compared to ZyXEL's
which performed better at 21.5K." Loop 6 also haunted some modems
mostly due to the fact that it defines a condition that is
equivalent to Loop 1 on half of the circuit."(Briere p.42)

"ZyXEL Modem Did Perform As Advertised"
In the third test because of clear (perfect) line all modems were
configured to perform at their absolute optimal rate.  The DTE rate
was set to the maximum allowed with each modem.  Data compression
and error correction were turned on, while the line rate was set to
auto and terminal flow control was RTS/CTS.  Five asynchronous
files were went over a clear line.  Here, ZyXEL performed well at
its 76.8K bit/ sec speed. 

"ZyXEL was a contender in nearly every throughput race, finishing
second in four out of five test."......Front-panel displays were
obviously critical in our tests, and there is a remarkably
different look and feel to each one.  The display on ZyXEL modem
seemed the most user friendly." (Briere p.41) 

Source:  Briere, Daniel,. "Test underscores differences in modem
throughput. "NETWORK WORLD May 10,'93. 





