[SI-LIST] Re: Which tool is the best

  • From: "Michael E. Vrbanac" <vrbanacm@xxxxxxxxxx>
  • To: "Michael Khusid" <mkhusid@xxxxxxxxxx>
  • Date: Sat, 17 Apr 2004 09:58:25 -0500


I didn't know simulation products needed "calibration".  What would you 
suggest the
calibration schedule to be?  About once a year... say about the time for 
paying for the
maintenance contract?  I was also wondering if you sold maintenance 
contracts on
"brain calibrations".  I'll probably need one soon. I hope my model isn't 
too old.
I also hope it won't cost too much because I spent too much on my 
simulators. <grin>

The brain is an amazing thing... and it loses its edge only if its owner 
wants it to and in
a long-term sense as a result of aging.  After years of success, I can only 
say that being
"right" is no guarantee for keeping one's job. Neither do simulators nor 
proof insures it either.
Sometimes its the opposite... you can be extremely good and folks hate you 
even if you're
friendly and helpful... its funny how that happens.  I know because I've 
been there, done that.
Nice try, anyway....

Anyway, it looks like I struck a nerve with you.  I was only having a 
little fun with the subject
and encouraging folks to learn from the tools they work with and test their 
knowledge by
"thinking outside the simulation box".  Most problems I solve are the 
result of "restricted
and programmed" perspectives with regards to engineering design (even about 
so the message was a light-hearted way of looking at that problem and was 
intended to be a
service to everyone.  This "simulation mantra" everyone seems to be 
chanting is a little
pathetic and a little short on logic and leads some folks to make some 
wrong decisions
about them.  Some time ago I saw a company where the common thinking wasn't 
out well so it doesn't always work out as rosy as some of the pictures 
painted for simulation.

I hope you didn't mean to say we all shouldn't use our brains, though.  New 
outlooks on
engineering doesn't require us to change physics.  And we all learn from 
doing the
simulations as well which might actually mean that in the long run fewer 
simulators will
be sold.  Strange but true.  Who will buy them after everyone knows what's 
going to

The truth about simulators is simple. They are only as good as those who 
drive them.
They are tools and nothing more.  Even as tools, they aren't perfect. I 
know. I own several
(and like them) and have worked with a number of others as well and as a 
user of simulation
tools I am aware of what they will do and what they won't. As such, I will 
share my own
"The 15 'Immutable' Laws of Simulators" ... perhaps a few will make you 

1.  The Simulator Law of User Omniscience
Simulators work best for the user that already knows the answer before its 
and knows what they should expect as a result of one action or another 

2.  The Simulator Law of User Ability
A simulator's ability to perform well is directly related to the expertise 
of the operator.

3.  The Simulator Law of Garbage In-Garbage Out
A simulator's ability to perform well is directly related to the quality of 
the information
placed into it.

4.  The Simulator Law of Data Glut
Simulators can provide you with more data than you really need, correct or 
not, as fast
as any other method known.

5.  The Simulator Law of Limited Absolute Accuracy
Simulators can give the impression of great accuracy even though the input 
data has a wide
latitude for error.

6.  The Simulator Law of Limited Responsibility
The performance the vendor had stated for the simulator is not guaranteed 
in your case.
Any inability of the simulator is your fault.

7.  The Simulator Law of Restricted Application
Simulators do not convey any apparent "veracity" in anything other than the 
exact situation
they seemed to solve.

8.  The Simulator Law of Apparent Diligence
Simulators provide an excellent means to make you look busy for weeks even 
though you could
have come to the same decision in ten seconds by using your brain.  (Great 
to use at layoff time.
No, not the brain... that'll get you laid off.  The simulator....)

9.  The Simulator Law of Apparent Genius
Simulators provides the appearance that the operator is wise and 
all-knowing.  This can happen even
though the operator couldn't otherwise remember his own name...

10.  The Simulator Law of Data Quality
The degree of data quality of a simulator is directly related to its cost. 
A more expensive simulator
is obviously better and highly accurate simulators also must have high 
maintenance costs.

11.   The Simulator Law of Invincibility
Simulators provide job protection, cause all your decisions to be correct, 
make management like
you and give you a raise, stops hair loss, and improves your love life. (I 
had only 14 laws until you
helped me write this one.  Thanks!) (no that's not guaranteed... your 
mileage may vary; see #6).

12.  The Simulator Law of Altered Reality
As long as the simulator results look remotely similar to the time domain 
results, the results are
deemed accurate and are lauded industry-wide.

13.  The Simulator Law of Never-Ending Maintenance
The simulator has to work and you need the undying commitment of the 
simulator company
to help when things go wrong. You will not get this without maintenance. 
Expect to go broke
trying not to make mistakes.  Imagine the horror of being without 
maintenance. You are
considered lower than dirt and become instantly stupid.

14.  The Simulator Law of Infallibility
A simulator is never wrong.  If it is ever wrong, its a software bug and 
will be fixed in the next
maintenance release... that is, if you paid your maintenance fees...

15.  The Simulator Law of the Missing Vendor
The simulator vendor disappears if the software bug is too embarrasing 
hoping that you'll
become stupid and ineffective by not being able to use the simulator and 
thereby get laid
off and that'll keep things quiet. (Yes, and your hair falls out , etc.)

So the luckiest folks in the world must be those who work at the simulator 
vendors.  They
alone enjoy all the fruits of these "laws" without all the extra expense 
and don't suffer any
of the "downside".  All the rest of us poor folks can only hope we, too, 
can have a simulator
one day.  Hey, my dog wants one, too!  She wants to become "human"....

There... now you can send the "simulator thought police" to shut me up.  :-)


Michael E. Vrbanac

At 02:58 PM 4/16/2004, you wrote:
>Every tool needs "calibration". This applies to measurement equipment,
>simulation software and "common sense" aka engineering brain.
>Let's do some failure analysis. Imagine a failed design. An engineer who
>relied exclusively on his or her brain and "common sense" is shown the door
>immediately. If an engineer has some design justification in terms of
>non-predictive measurements or simulations, then there is a chance for
>engineer to fix the problems and re-spin a design.
>So, the best brains can lose calibration if they are not coupled with
>continuous practical experience. The fastest way to gain experience is
>through failures. The cheapest way to find failures is through software --
>you don't need to design and build anything, and you don't need any
>measurement equipment. So, my vote goes for a painter with a brush, or a
>signal integrity engineer with software tools.
>P.S. I would like to comment on the "which tool is the best" discussion, but
>as a tool vendor, I will resist.
>Michael Khusid
>Ansoft Corporation
>SI/HF Application Engineer
>25 Burlington Mall Road, 5th floor
>Burlington, MA 01803-4100
>Tel 781-229-8900 Ext. 134
>Fax 781-229-8624
>---------------------http://www.ansoft.com ---------------------
>-----Original Message-----
>From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On
>Behalf Of Michael E. Vrbanac
>Sent: Thursday, April 15, 2004 7:09 PM
>To: fzanella@xxxxxxxxxxxx
>Cc: si-list@xxxxxxxxxxxxx
>Subject: [SI-LIST] Re: [SI-LIST]: Which tool is the best
>Well said.
>Also when we talk about which tool is best, have we thought about
>effectiveness vs. cost?  Its probably also important to mention that these
>tools are not "free", very focused on their utility (a nice way to say
>and are not always necessary (although I am sure there are folks who'll
>with me on that one).  We like to argue about how much we can save on a
>program if we simulate.  The counter to this how much can we save if we
>need to simulate?  And while its fine and good to show how many board turns
>you can save by simulating, it may be just as good to ask, how many board
>could you save if everyone used plain ol' common sense before they turned on
>the simulator?  Those of you with a little gray showing in your hair know
>I'm talking about.
>Now all those tools that have been mentioned are really good and I'm
>experienced at a few of them but there is no substitute for a clever
>using his brain instead of his pride (or something else) to get his designs
>done well.
>And a poor design decision is still poor even if the person has the best
>in the world at his disposal.
>What's your salary per year?  How much do these tools cost per year
>Do you know you are your company's best bargain in that comparison?  Its is
>easy to outstrip what you are paid by the yearly cost of the simulators you
>might run.
>So the issue comes down to "how many decisions you can make without the
>measures your effectiveness in that comparison. Not only that, engineers
>are running
>around meeting that tight schedule making critical decisions while the bulk
>of the time,
>that simulator that sits out there on the computer still costs that
>maintenance fee even
>while it "sits" around doing nothing.  That might remind you of some
>co-workers you've
>worked alongside of, doesn't it.
>I have often seen the paralysis that comes upon engineering teams these
>We cannot even make one decision without "we've got to simulate it".  While
>certainly some situations do require such caution, others clearly do not.
>But humans can mistakes, you say.  Indeed, but one bad model file can render
>the world's best simulator worse than ineffective. It can be downright
>If one doesn't know better, it can actually make you think that you have a
>dependable answer and spend millions upon millions on a bad idea.  So if you
>don't plan to use your brain, don't get any simulator.  Your company will
>thank you
>while they are walking you out the door.
>My vote is clear.  Use your brain.  Its cheaper, faster, really handy (its
>you know) and doesn't require a rechargable power source.  So don't leave
>without it.  Now if you have the opportunity to couple it with some of
>these fine tools
>and can justify the cost, that sounds like a winning combination to me.
>Best Regards,
>Michael E. Vrbanac
>At 02:50 PM 4/15/2004, you wrote:
> >I agree with Arpad's comment.  The best approach is to have multiple
> >tools at your disposal.  There isn't one tool which can address all
> >requirements.  Of the tools available today, a couple can be classified
> >as being very good at SI simulations on boards - reflections, crosstalk,
> >delays, I'll call these Class A tools.  There are a couple of different
> >tools which are good at simulating power integrity, return paths; these
> >Class B tools cannot compete with the Class A tools at SI simulations.
> >Then there are Class C tools whose strength lies in high frequency (2.5
> >GHz and above) analysis, using 3D solvers.  And let's not forget timing
> >analysis, only one tool, mentioned in a previous email, performs board
> >level SI and timing analysis.
> >
> >Fabrizio Zanella
> >
> >-----Original Message-----
> >From: Muranyi, Arpad [mailto:arpad.muranyi@xxxxxxxxx]=20
> >Sent: Thursday, April 15, 2004 2:18 PM
> >To: si-list@xxxxxxxxxxxxx
> >Subject: [SI-LIST] Re: Which tool is the best
> >
> >A slightly different "opinion":
> >
> >You also have to know how to ask a question to get meaningful
> >answers.  The original question was something like which tool
> >is the "BEST".  Pretty soon we got a list of practically all
> >available tools on the horizon.  So which is THE BEST if all
> >are listed?
> >
> >Of course, the definition of BEST is relative, it all depends
> >on what you want to do with the tool.  One may be excellent
> >for one thing while completely useless for another thing...
> >
> >Arpad Muranyi
> >Intel Corporation
>To unsubscribe from si-list:
>si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
>or to administer your membership from a web page, go to:
>For help:
>si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>List FAQ wiki page is located at:
>                 http://si-list.org/wiki/wiki.pl?Si-List_FAQ
>List technical documents are available at:
>                 http://www.si-list.org
>List archives are viewable at:
>                 http://www.freelists.org/archives/si-list
>or at our remote archives:
>                 http://groups.yahoo.com/group/si-list/messages
>Old (prior to June 6, 2001) list archives are viewable at:
>                 http://www.qsl.net/wb6tpu

To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List FAQ wiki page is located at:

List technical documents are available at:

List archives are viewable at:     
or at our remote archives:
Old (prior to June 6, 2001) list archives are viewable at:

Other related posts: