From: Scott McMorrow ([email protected])
Date: Mon Dec 18 2000 - 10:28:32 PST
Actually, the problem is much simpler than you make it.
The problem lays with the silicon vendor and the companies which
use their silicon. If you "need" simulation models in order to
perform a good engineering design ... and you cannot obtain
them from the vendor of the silicon ... then you should put pressure
on them to provide, or contract out the model development.
As the purchaser of the device, you have the only leverage available.
If designers start making analog simulation models a part of the
contractural obligation for the purchase of silicon, then this problem
would go away rather quickly.
Can anyone imagine performing an ASIC design, synthesis and
verification without having accurate models and specifications of
the cell libraries?
We execute 20 to 40 SI evaluations for customers each year and
have the same issues as you. When our customers are unable
to obtain the data necessary to perform accurate analysis and
design, we counsel them to:
1) place leverage on the vendor to obtain the data
2) switch vendors
3) add ibis/spice modeling to purchasing contracts.
If you can't obtain the information that you need from Toshiba,
what makes you think that a consulting house or a software
Ultimately, if you need a particular piece of data before executing
a design with a specific part and are unable to acquire it, then your
design is at risk. I often ask my customers to change parts in
their designs for this reason alone.
-- Scott McMorrow Principal Engineer SiQual, Signal Quality Engineering 18735 SW Boones Ferry Road Tualatin, OR 97062-3090 (503) 885-1231 http://www.siqual.com
Ken Cantrell wrote:
> Scott, > "If you have a problem with a particular vendor's approach, then by all > means boycott that vendor and do it yourself. But when you "do it yourself" > please remember to figure out what the true fully burdened cost of your > engineering work truly is. You will find that you actually spend much more > in real accounting dollars than you would have spent with an outside > source, and in many cases your result will not be as good as theirs. They > have the time to develop a "quality" process." > > Just a clarification. > We are all unfortunately familiar with an "IBIS ain't an IBIS" sometimes; it > is frequently discussed right here. The problem for the majority of us is > that the trail ends at not being able to get the files, any files. Then we > are stuck with generic driver simulation, which is better than nothing, but > not much. Installing a hardware lab to perform IBIS compliant testing is > cost prohibitive for the majority, and I agree, would soak up the > engineering time. Mentor doesn't have a hardware lab either. We need > someone that has a lab .AND. a dedicated "software" group, integrated and > focused to produce truly compliant spice/ibis files in a pre-release mode. > Pre-release. Bold that last sentence, stressing pre-release, and forward it > upstream to as senior a marketing person as you can. Title it "Marketing > Direction Imperative". Partner with an outside lab or grow your own. I > really should be charging you guys. > Another alternative in lieu of a lab, but it would be better if you had > both, is to have a group, a subset of what Mentor has now, dedicated to > obtaining this information from the vendor (we know about MAP, right idea, > just not enough). Lobby the suckers, be forceful, be the voice of the > consumer base. Sleeping giants are hard to wake up. Our need for > pre-release files is growing constantly ( or exponentially it seems). Once > you have the files, it is trivial to produce the final product even > factoring in burdened engineering dollars. Mentor, Cadence, Innoveda, et > al, the one-stop shopping giants of this industry, should shift focus. The > problem is getting the information from the vendors, not processing it. The > thing inside the box is the data, the gift wrapping is the spice/ibis file. > > I concur that economies of scale exist when a group focuses on a single > product versus "do it yourself". Mom & Pop shops can't compete in > quantity/cost, and sometimes even in quality, depending on how the large > operation is run. My intent here was not to slay another dragon, but to let > you know that the market direction is incorrect. It is not supplying what > the customer base needs. There's money to be made here because the demand > will only increase with time. The company that delivers the right product > at the right price will take the lion's share of the market. And don't > forget to unbundle it from you product line so you don't look like a shmuck. > > Ken > -----Original Message----- > From: Scott McMorrow [mailto:[email protected]] > Sent: Friday, December 15, 2000 10:00 PM > To: Ken Cantrell > Cc: Dunbar, Tony; [email protected] > Subject: Re: [SI-LIST] : searching for Toshiba models > > All, > > If the process of creating a high quality IBIS model from the vendor's > HSPICE deck and s2ibis were all that were necessary, then > Ken's statements would be quite true. Unfortunately, this is not the > case for many instances. > > There are several vendors out here who perform the service of IBIS > model generation and correlation. For some the cost is in bundling the > service with a particular tool. For others, like ourselves, it is charging > the > true engineering cost of creating and validating these models. > > The IBIS specification is free to all who would choose to use it. But > anyone > who has participated in the specification compliance process knows that > it is one thing to say than one complies with a specification, and quite > another > to actually comply. > > If you have a problem with a particular vendor's approach, then by all > means boycott that vendor and do it yourself. But when you "do it yourself" > please remember to figure out what the true fully burdened cost of your > engineering work truly is. You will find that you actually spend much more > in real accounting dollars than you would have spent with an outside > source, and in many cases your result will not be as good as theirs. They > have the time to develop a "quality" process. > > best regards, > > scott > > -- > Scott McMorrow > Principal Engineer > SiQual, Signal Quality Engineering > 18735 SW Boones Ferry Road > Tualatin, OR 97062-3090 > (503) 885-1231 > http://www.siqual.com > > Ken Cantrell wrote: > > > Tony & all, > > This is a long one (all), but well worth the wait. Please read to the > end. > > > > You'll have to forgive my "forceful posting", it's a 60's thing. If there > > were another avenue I would certainly use it. It is a serious problem, > and > > the only way you can get some movement, after other means have failed, is > to > > be forceful. Sometimes a wilderness adventure does the trick. > > > > I vectored to your site via your link and looked at MAP and qualityIBIS. > > Data referenced was cut and pasted from you own site. > > > > qualityIBIS: > > "Visual inspection for completeness and reality > > Model checking using ibis_chk > > Graphical inspection using s2iplt or winibis > > Simulation using ICX " > > > > *s2iplt is available at eda.org for free. Perl is required to run the > > application. It is free. WinIBIS is posted on the IBIS home page and is > > supplied by HyperLynx, for free. Everything here except simulation with > ICX > > is public domain. Anyone can sim the file with the sim pack of THEIR > > choice. > > > > ICX Standard Library > > 11,000 models > > Licensed per site > > Low cost and regular updates. > > Updated every six months. > > New models added with each update. > > *Doesn't Cadence do the same thing (same amount of files), just bundled > with > > the product? > > > > ICX Project Modeling > > "From the parts list supplied to us at the time of order we provide > > qualityIBIS models from our Standard Library matching the exact devices > you > > are using or create models from devices or a SPICE source you supply." > > *Why don't you throw in the Standard Library along with the Custom models, > > it's cheap. > > *Created from devices. Your group doesn't create the files from devices, > > the user has to supply the information themselves from their lab data. If > I > > had a lab set up to measure and collect all of the data required for IBIS > > compliant testing, I would certainly have the capability to create the > IBIS > > file. Second case, user supplied SPICE file. If I have the SPICE file, I > > can convert it myself with free spice2ibis converters in a selection of > > OS's. At: http://www.eigroup.org/ibis/tools.htm. > > > > The marketing plan and goal for the project can be found at: > > http://www.mentor.com/icx/icx_models/alliance.html. It's a good concept > and > > it's nice that Mentor has taken the first step. A small incremental cost > for > > the convenience of having all of the IBIS files in one place is worth > > paying, especially for people who don't know where all this stuff is > > located, and really don't care. They just want the files. What I object > to > > is your tying the service to exclusive use of Mentor product (sold at > > reduced rates as per the above url). It's kind of like the > > Microsoft/Netscape thing, you know? Bob Ross, from Mentor, IBIS Open > Forum > > Librarian, vectors anyone, for free, to all available IBIS sites, indexed > by > > vendor. Excellent work Bob. Many thanks on many occasions. Done in the > > true spirit of the IBIS group. This is at /ibis/model.htm instead of > > /ibis/tools.htm. > > > > So far you aren't capable of providing any service that we can't provide > for > > ourselves given access to the same information, the lab data or the Spice > > files. > > > > MAP: > > Below is the member list of the alliance. You will note, Tony, Toshiba > > isn't on there. Remember, this was all about not being able to get a > > Toshiba part? Excerpt from your e-mail: > > > > Anyone who finds themselves in Christoph's situation where a silicon > vendor > > cannot, for whatever reason, provide IBIS models directly may want to > > suggest to them (or encourage them) to look into the MAP. > > > > *Why should we vector them to Mentor? You are trying to lock us into an > > exclusive contract. I prefer free choice, free speech and free thought, > > thank you very much. > > > > In conclusion, I think Mentor needs to rethink it's marketing strategy for > > the group. It would be whiz-bang if you had an on-site lab that performed > > the IBIS testing, and pulled it into both spice and ibis formats. Now > that > > would be worth buying, and you would be providing a service that is very > > much needed, and in line with the Open Forum intent. I would also suggest > > that you change the "buy the Mentor product or no deal" mode that you have > > now, and let your mainline product stand on it's own merits. I hear it's > > pretty good, but I don't like to be forced into doing anything. The > > projected marketing pull-through won't happen if you are perceived as > > monopolistic. > > I will be reviewing your product in Q1. I have the Tau and Interconnectix > > disks, and will request a license file upgrade soon (ran out of time on > the > > first one, too busy to even open the disks). Maintaining the intent that > > Bob Ross and the other founders of the forum had, as an aid to Signal > > Integrity, is more important than your current marketing direction. Or > to > > put it another way, what would you think if HyperLynx started charging you > > to use WinIBIS? You are not providing a true value add at this point, you > > are simply shuffling paper. > > > > Tellin' it like it is, > > > > Ken > > > > Post script: (All) Todd Westerhoff (uses SpectaQuest as I remember) sent > an > > e-mail stating that Toshiba supplied him with HSpice and IBIS for one > their > > products, but makes no claims to broad product line support. > > ******* > > *Member list of MAP* > > > > Membership > > > > Actel Corporation > > > > Advanced Micro Devices, Inc. > > > > Agilent Technologies Inc. > > > > Cypress Semiconductor Inc. > > > > Dallas Semiconductor > > > > Data Device Corporation > > > > Micron Technology > > > > Mitel Corporation > > > > National Semiconductor > > > > Pericom > > > > Sharp > > > > Ken > > > > -----Original Message----- > > From: [email protected] > > [mailto:[email protected]]On Behalf Of Dunbar, Tony > > Sent: Friday, December 15, 2000 11:05 AM > > To: '[email protected]' > > Subject: RE: [SI-LIST] : searching for Toshiba models > > > > Hi all, > > > > Michael Nudelman is quite correct, of course. Our modeling group probably > > could do these IBIS models either from measurement or Spice. Yes, there is > a > > fixed-price fee and it probably equates to about the cost to your > company's > > overheads for you spending about 3 hours trying to figure out how you're > > going to get an IBIS model for these parts or fix-up what you have! > > > > I have also read Ken Cantrell's forceful posting on this issue, too. There > > is another possible approach here which does not involve banishing any > > vendor out into wilderness: > > > > Part of Mentor's IBIS modeling strategy and solutions is the Model > Alliance > > Program. I have included a URL below where more introductory information > can > > be found but, essentially, MAP is based on a partnership between Mentor's > > IBIS modeling group and silicon vendors whereby Mentor's modeling group do > > the IBIS modeling directly for the silicon vendor. The models are then > > provided back to the silicon vendor for him to distribute as he wishes; > e.g. > > posting them on their web-site for all to download. > > > > Anyone who finds themselves in Christoph's situation where a silicon > vendor > > cannot, for whatever reason, provide IBIS models directly may want to > > suggest to them (or encourage them) to look into the MAP. > > > > Here's the URL: http://www.mentor.com/icx/icx_models/index.html > > > > Anyone with specific interests or questions, including silicon vendors, > > please contact me directly off-line from the SI-LIST. > > > > Regards, > > Tony Dunbar > > Mentor Graphics/ICX > > Dallas, TX > > > > -----Original Message----- > > From: Michael Nudelman [mailto:[email protected]] > > Sent: Friday, December 15, 2000 8:06 AM > > To: Christoph Hillen > > Cc: [email protected] > > Subject: Re: [SI-LIST] : searching for Toshiba models > > > > I never tried it, > > > > But Mentor has division that will take your part and create IBIS model of > > it. > > Probably, for a fee. > > > > Christoph Hillen wrote: > > > > > hi all, > > > > > > I'm searching for IBIS models of the Toshiba devices > > > > > > TH58V128FT > > > TC58128FT > > > TC55V4000ST-70 > > > > > > It seems that the company itself does not provide models, and they are > > obviously > > > not able to answer any questions regarding models. > > > Does anyone of you know a source for these? > > > > > > Thanks a lot > > > > > > Christoph Hillen > > > Utimaco Safeware AG > > > Germany > > > > > > **** To unsubscribe from si-list or si-list-digest: send e-mail to > > > [email protected] In the BODY of message put: UNSUBSCRIBE > > > si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP. > > > si-list archives are accessible at http://www.qsl.net/wb6tpu > > > **** > > > > **** To unsubscribe from si-list or si-list-digest: send e-mail to > > [email protected] In the BODY of message put: UNSUBSCRIBE > > si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP. > > si-list archives are accessible at http://www.qsl.net/wb6tpu > > **** > > > > **** To unsubscribe from si-list or si-list-digest: send e-mail to > > [email protected] In the BODY of message put: UNSUBSCRIBE > > si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP. > > si-list archives are accessible at http://www.qsl.net/wb6tpu > > **** > > > > **** To unsubscribe from si-list or si-list-digest: send e-mail to > > [email protected] In the BODY of message put: UNSUBSCRIBE > > si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP. > > si-list archives are accessible at http://www.qsl.net/wb6tpu > > ****
-- Scott McMorrow Principal Engineer SiQual, Signal Quality Engineering 18735 SW Boones Ferry Road Tualatin, OR 97062-3090 (503) 885-1231 http://www.siqual.com
**** To unsubscribe from si-list or si-list-digest: send e-mail to [email protected] In the BODY of message put: UNSUBSCRIBE si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu ****
This archive was generated by hypermail 2b29 : Tue May 08 2001 - 14:30:27 PDT