News Archive

Well, I've changed jobs. I'm now working at Motorola, although I am . My new email address is . (1/16/99)

There is an interesting interview with Mark Law on the Interview page. (1/16/98)

of a Win32 port of PISCES-IIB 9009. (1/8/99)

A reader points out that I missed a - 150K shares in August. (1/8/99).

First announcement for CHiPPS 2000. (1/6/99).

from one of my alma maters. One of my favorites is a. I am having better luck with some of these using IE than Netscape, but that might be specific to my site. (1/5/99)

More on implant table discontinuities - the case of the tunneling implant. (1/5/99)

The January issue of Electron Devices contains a pair of interesting simulation papers: Three-Dimensional DIBL for STI MOSFET's (p. 139) and Design Considerations of High-k Gate Dielectrics for Sub 0.1 um MOSFET's (p. 261). (1/5/99)

Steve Bourland's Master's thesis, Device Benchmarking of MOSFET Device Simulators is available by request. (1/5/99)

I have added an excellent contributed IEDM summary to the Calendar page. (1/5/99)

Avant!/Cadence updates will now be reported in a separate section. See left column. (1/5/99)

It finally dawned on me that I could use a parallel version of Atlas after all. For large splits, you are better off running the conventional version, but for "debugging mode", the parallel version can help a lot, even on a 2 CPU machine. (12/18/98)

I'm looking for inputs/suggestions on why the TCAD Journal isn't more popular. Also, does anyone have any summaries/comments on IEDM papers? (12/18/98)

I dug up some of my old posts to the . The first is a conversation on in PISCES. Improvements have been made since then, but it's still interesting (IMHO:). Here is an update on that topic from Andy:

These comments were written in 1992 before I had any chance to view the operation and effectiveness of the Stanford CURVETRACE routine. Since that time Silvaco has implemented, developed and recommended the use of this algorithm for users wishing to perform complex IV curve extractions. Given appropriate inputs there are many cases where the machine controlled bias stepping is able to extract smooth results of curves such as second breakdown and CMOS latchup. Of course there are still cases where the automatic bias stepping fails and user control with compliances as described here is more efficient. The second article contains some You can find these same points elsewhere at this site, but this is a nice concise summary. (12/17/98)

Need some help with your simulations? that offers to help. (12/8/98)

Avant! announces the . This is good news, although it's too early to be calling it a "standard" (12/8/98).

TCAD developers: interested in getting your software ported to Linux with the help of Redhat? (Dec 2, 1998).

Let's see - shares should be about $460K (Nov 24, 1998).

I know of a job in Sunnyvale for an experienced TCAD engineer, if anyone is interested. (Nov 17, 1998).

Here is the new link to (Nov 5, 1998)

. Particularly interesting is the comment ``we have been dissatisfied with cost effectiveness, particularly in the services business" because this is the direction TCAD vendors have been headed. (Nov 4, 1998).

A story about . It should either explain why salesmen are the way they are, or confirm your worst fears about salesmen, or both! From Wired magazine. It's long, but I recommend it. Personally, I still wonder why we have non-technical salespeople in this industry. (Nov 2, 1998)

When I worked at TMA, "AURORA" was the name of the SPICE-parameter extractor. Now, it is the name of the Avant! Users Group. They are meeting in February, but be aware that this is for all Avant! products, not just TCAD. (Nov 2, 1998).

If you haven't already made your IEDM plane reservations, you might be interested in this 2D dopant profiling meeting. (Nov 2, 1998).

In addition to the above two meetings, there are lots of new links in the Calendar (Nov 2, 1998).

I may have a Win32 port of PISCES 9009 available for you soon. Stay tuned. (Nov 2, 1998).

Just added a TMA-specific Tips and Tricks page, courtesy of Carl Huster. I was tempted to call this Carl's Corner, after the truck stop of the same name that's somewhere between Austin and Dallas. Note that tips from one vendor will occassionally apply to other vendors as well. (Nov 2, 1998).

Some follow-up Q&A to last issue's device benchmarking report by Steve Bourland. (Nov 2, 1998).

There are a couple of new papers in the One is on halo implants, but be careful, it's a big file that choked on my PC, although it works fine on my unix machine. The other is on boron segregation - this was previously presented at SISPAD in Boston, but there are more details now. (Nov 2, 1998).

Here is a quote that should scare the vendors: in the October issue of Wired, page 97, Symantec lawyer Derek Witte worries that the Y2K bug "could infect the entire industry with the virus of litigation. 'The inference is that any software bug gives rise to product liability.'" (Oct 12, 1998).

Here's a first call for papers for a TCAD Applications conference. Actually, the conference doesn't exist yet. The first step is determining whether there are enough people who would be willing to submit something and/or travel to such a meeting. The emphasis would be on how TCAD tools get used, and how they have helped you, and not on how another new model shows good agreement with measurements. Ideas on this welcome. (Oct 12, 1998).

What I Did This Summer, by . It's mainly a benchmarking study of several device simulators. There are surprisingly large differences between vendors on allegedly the same problem. I placed this in the calibration section. (10/12/98).

It's been a while since I added anything to the Calibration page. There are two new contributions, courtesy of Micron. One is an example of implant table discontinuities, and the other are some alternate implant moments generated with UT-MARLOWE (Oct 5, 1998).

reports that he has ported Stanford's PISCES 9009 to Linux (Oct 5, 1998).

Device Designer is now available from . This is a PC based MOSFET simulator with experimental spreadsheet. (9/16/98).

I have been negligent in not mentioning the web based simulation interface at Purdue: . Avant! (TMA) is the only vendor participating so far, but lots of university programs are available for a test drive. (9/16/98)

But that doesn't include TCAD. See next story. (9/16/98).

An interview with Mark Pinto, CTO of Lucent's Microelectronic's Group, on the future of TCAD (9/16/98).

Dave Clark writes that he has Rel98 of FLOOxS working on his Pentium box with Red Hat 5.0. if you are interested. (9/14/98)

I previously told you that P.K. Mozumder has joined PDF Solutions. You may have assumed like I did that he was in San Jose. Instead, he is in their new Dallas office. (9/14/98)

I have written a book review of Semiconductor Device Physics and Simulation by Yuan and Liou that should appear in IEEE Circuits and Devices. Bottom line: I don't recommend it. (9/14/98)

There is another new simulation book out there, Semiconductor Devices, A Simulation Approach, which I haven't seen yet, but there is an It also come with a general purpose PDE solver, and one of the authors writes that it is easy enough to use that undergrads can numerically solve EM or diffusion problems with just a few minutes coaching! (9/16/98).


You may have heard that the original judge in the Cadence/Avant! case has because he talked to a reporter about the case. More recently, there was a (Sorry, this link has expired). (8/2/98)

stands for Software and Analysis of Advanced Materials Processing Center. It's the University of Florida TCAD site. (7/9/98) The 2nd annual will be held Nov 2nd (7/16/98). As long as Stanford is no longer holding annual reviews, you should consider attending this instead (9/16/98).

is available at this link. It's a free ion implantation modeling program. It's a classic program, now a subset of SRIM. (7/9/98)

More links: the , and a TCAD success story from (7/9/98)

Here is the link to the call for papers for the . (7/8/98)

Cable modems are coming to Austin! What does this mean? Well, there is a good chance I will have to switch ISP's to get the new service, so be sure to get added to my to find out when/if I am moving. (7/7/98)

There are a pair of new TCAD papers that are worth checking out. IBM has one in the July Electron Device Letters (p. 234) about BF2 diffusing along the silicon/oxide interface. A rare example of TCAD being used to predict something before it is physically measured. Also, the February issue of the Trans. on CAD, page 149, has a paper on grid quality. (7/7/98)

There were several papers of interest to TCAD'ers at the Statistical Metrology workshop. If you know someone who attended the VLSI Symposium, the papers are on that CD-ROM. (7/7/98)

I have added some more information on the next Ultra Shallow Junctions Workshop (6/29/98)

Two new questions in the FAQ. Why are obtuse triangles bad, and what about phos and arsenic co-diffusion effects? (6/23/98)

P.K. Mozumder rejoins his old office mates by moving to PDF Solutions. (6/23/98)

Here's a paper from the recent Statistical Metrology workshop where we used a process-to-device grid translator to look at grid sensitivities. The amazing thing here is that I was able to together. (6/23/98)

Xing Zhou has a nice web site. I like his project (Design and Optimization of Ultra Small Transistors, and a of a CMOS process. (6/23/98)

There will be a special issue on TCAD in the June 99 issue of the IEICE Transactions on Electronics (produced in Japan). The call for papers is available in the May issue. (6/23/98)

The SRC is planning to conduct a 2-day workshop in early August to develop a long-range strategy for Modeling & Simulation for the SRC. The preliminary title of the workshop is "Semiconductor Modeling Beyond the Continuum: Towards an Atomic-Scale Digital Windtunnel". Contact for more details.

IEDM will have a TCAD short course this year. (6/22/98)

I've added my favorite undocumented TCAD trick to Tips and Tricks. 5/12/98

Some work we did here on metal line RC extractions using SUPREM4/PISCES added to Publications. 5/12/98

Interesting new TCAD survey at . Look under , then select Survey from the left column. So far, just interconnect modeling, but there are placeholders for other areas in TCAD. 5/12/98

"Award-winning" device simulation software with optoelectronic capabilities (lasers) available from . 5/12/98

Valery's new company is called . 4/27/98

What's new in CMP modeling? You need to ask , ex-MIT, now at , or , at the Southwest Research Institute. (4/21/98)

2D dopant profiling has been a highly ranked need in the National Roadmap (NTRS). But the desired spatial resolution may be beyond what is physically possible. So what do we need instead? Here is a first pass; further comments are welcome: (4/21/98)

    Dopant Profiling Needs
    Lateral junction location at silicon surface. 
    Distance from surface junction to gate edge (overlap). 
    2D channel profiles (channel, not S/D). 
    2D point-defect profiles. 
    1D surface profiling (dose loss). 
    1D active concentration at bottom of poly.

ISE is greatly expanding their U.S. presence, with a new office just opened in San Jose. The CEO will be located there. (4/21/98).

Valery Axelrad is no longer with PDF Solutions. Perhaps you weren't aware that he left TMA last year. Stay tuned for further updates. (4/21/98).

So now Digital Instruments is going to merge with Veeco (4/21/98).

The National Roadmap has some rather aggressive 2D dopant metrology specs. So aggressive, they might be impossible. I re-examine what the dopant metrology needs really are, and explain why they are so difficult to measure, to help arrive at a better compromise (3/19/98).

Speaking of roadmaps, a is now available. This is a nice, detailed list of some of our future modeling needs (3/19/98).

The National Roadmap is out. Although the Roadmap is intended to be a list of "technical" needs, the more serious limitations to TCAD's progress are related to cooperation, tech transfer, setting proper expectations, etc. Here is my version of TCAD needs (3/9/98).

USJ-99 will be held March 28-31 in Research Triangle Park, North Carolina. The full title of this conference is the Fifth International Workshop on the Measurement, Characterization and Modeling of Ultra-Shallow Doping Profiles in Semiconductors. "Modeling" is a new addition to the title. Historically, this workshop has been dominated by SIMS/SRP people, but TCAD'ers are welcome. Abstracts due by Sept 15. Hope to see some modeling papers there (2/24/98).

Many of the papers from USJ-97 were just published in JVST-B, Jan/Feb 98. Here are the foils from my presentation on shallow junction requirements (2/24/98).

There has been a link to the conference on my calendar, but what I didn't tell you was how many 2D dopant profiling papers there were. You can browse the now (2/24/98).


Here's a link for the (2/24/98).

Time to hire some more lawyers. MicroUnity is suing Precim for patent infringement (Solid State Technology, Feb 98). Recall that TMA purchased Precim and then later merged with Avant! (2/17/98).

I was recently asked if I had ever heard of level-set methods. Sure, I even have a link here on that, or so I thought. Turns out I never posted the highlights from last years' ChiPPS conference. Here they are:

That same week, an article appeared in the February Solid State Technology on topography modeling using level-set methods. There is also another article on ion implant modeling (2/17/98)

The "Year 2000" problem (Y2K) isn't really a TCAD issue (at least I don't think it is). But I liked this quote from an on this subject that relates to software engineering in general (2/17/98):

"But the result is that we have accumulated very little science or engineering expertise in intelligently evolving existing systems. We are fixated only on their initial creation as if that is all that matters. Whatever tools and techniques we have derive from that endeavor and they are ill-suited to the more difficult aspects of system evolution. Go to the Web site of most major universities' computer science departments and check the number of undergraduate and graduate courses available in the maintenance or evolution of existing systems. I could find none in a relatively brief search."

REED is a molecular dynamics (MD) ion implantation modeling program being developed at Los Alamos. Check out their . The software is only available to CRADA members, but you can see some of the results. MD is supposedly better than the traditional binary collision approximation (BCA) at low energies, but there is still some life left in UT-MARLOWE! Version 4.1 should be arriving soon, and they have made extensive changes that allow lower energy implants (2/17/98).

has crawled onto the web. Dave has (1/29/98).

Here is (1/29/98).

Well, TMA is now officially part of Avant!. First change is the There continue to be a bunch of press releases by Avant! and Cadence regarding their on-going lawsuit, but here is an interesting one you may have missed (1/29/98).

Looks like TMA is moving again. Their building has a "For Lease" sign up. (12/9/97)

will be leaving the SRC at year-end. Haven't heard who the replacement will be yet. (12/3/97)

The proposed merger between Zygo and Digital Instruments is off now. (12/3/97)

is retiring from TCAD at year-end. Congratulations, Mike, although I suspect we'll see you again. (11/24/97)

Very glad to hear about the co-operation between Silvaco and PDF Solutions - something this industry could use a lot more of. See their booths at IEDM. (11/24/97).

I was thinking of making an for people interested in knowing when there was significant changes to this site. Let me know. (11/24/97).

Interesting new web page on (CFD) modeling for semiconductors. In other words, equipment modeling. (11/24/97)

Stanford's Center for Integrated Systems (CIS) is

UTQUANT, a quasi-static CV simulator including quantum mechanical (QM) effects has been released. Contact to see if it is available in your area.

Maybe we need a new section just for lawsuit news. Anyway, Meta (now part of Avant!), has lost a to Silvaco. (11/10/97)

intends to acquire Digital Instruments by year end. The latter company is developing a tool for 2D dopant profiling. (11/11/97)

How will Cadence's remarks regarding play in Japan? (10/14/97)

More TMA news: are down, and in its Platinum Partner Connections program. (10/8/97)

You know about the TMA/Avant! merger. If not, read the article. This makes the Avant!/Cadence lawsuit more interesting to TCAD'ers. See the , starting around early September. There is also an interesting interview with Joe Costello on this topic in . (10/3/97)

is now with Reaction Design, a start-up equipment modeling consortium. (10/3/97)

was held in Germany in August. The purpose of this workshop was to bring together scientists with a wide range of expertise to foster new interactions and to prioritize modeling needs. You can contribute to this process via a . (10/3/97)

PROPHET - Lucent's process simulator - may be available soon on a workstation near you. Stay tuned. (10/3/97)

Home * Calendar * Calibration * Tips and Tricks * Publications and Presentations * Editorial/Humor * Miscellaneous

This page last updated Mar 15, 1999 by