The idea of creating a more in-depth test discipline is not a new one. We in the educational community have been talking about it for years, but progress has been slow. The reason is simple: The number of people who can cite employment in the test domain is not that large. Consequently, students are not terribly eager to specialize in test, especially undergraduates.
Also, a serious curriculum involving test requires a serious investment in test equipment, which is too expensive for universities to own. So, practically, there’s no workbench available for students to use.
It’s not just an issue of obtaining the testers. The industry is very eager to donate obsolete testers and has done so on several occasions. The difficulty lies in tester maintenance. It’s a big, sophisticated piece of equipment that essentially requires a large amount of space and a technician to be available 24/7. Universities don’t have these kinds of resources.
One valuable solution that does not currently exist would be something equivalent to the MOSIS consortium, funded by DARPA, which provides silicon capability for students. The MOSIS Educational Program provides free fabrication of ICs designed by students in organized classes or involved in research at accredited educational institutions. Fabrication of dedicated educational wafer lots and research designs currently is cosponsored by the Semiconductor Industry Association (SIA), the Semiconductor Research Corporation (SRC), AMI Semiconductor (AMIS), DuPont Photomasks, and MOSIS. SRC manages the funds from SIA, AMIS supplies wafers for their processes, DuPont provides discounted masks, and MOSIS contributes program administration. More information can be found at www.mosis.org. If this kind of consortium maintained the tester and provided access to the students from various universities, a more in-depth test-engineering discipline might be feasible.
A testing curriculum does exist, in some form or other, in most engineering programs. It’s important to understand that adding such a curriculum to universities is not a matter of making the decision to do so. It’s a matter of having the infrastructure and appropriate materials.
A textbook is a key component. A valuable textbook that I believe represents a big step toward such a curriculum is Essentials of Electronic Testing for Digital, Memory, and Mixed-Signal VLSI Circuits by Vishwani Agrawal and Mike Bushnell, Kluwer Academic Publishing, May 2002. But the fact remains that it’s simply not going to happen overnight.
There are three key considerations to bear in mind with respect to developing a university curriculum focused on testing:
- VLSI design classes offer the greatest opportunity for expanding awareness. These classes produce pieces of silicon that must be tested, another argument for testing services similar to the MOSIS model.
Expanding the very popular VLSI domain into testing essentially would force students to learn how testers work and the value of testing. Students will appreciate the complexity of the testing domain only if they are asked to write the tester software for their designs. Even then, only a fraction of students will become interested in the test domain per se.
Testing by itself is not going to be attractive to students because of the size of the job market, but testing as a service to confirm that their designs are correct is very attractive. This would encourage students to become more interested and aware of the entire process and be a big step toward creating an interdisciplinary model.
To get a good school to teach something, you have to convince students that it makes sense, and students at schools like Carnegie Mellon are very knowledgeable. In our Ph.D. program, we have students who are industry experts, and they make decisions very consciously. My opinion is that just preaching testing as an interesting domain is not good enough. - The focus should be on testing as an activity for manufacturing rather than on the testing hardware. Testing hardware is a very narrow, sophisticated domain, so teaching tester architecture or electronics as a discipline probably is not going to happen because of the number of jobs available in this domain.
But teaching testing as a manufacturing activity is attractive. It’s slowly happening and may happen faster if some effort is made by the industry to create a consortium such as the one I’ve mentioned. It’s important for universities to focus on what will give students the greatest opportunity in terms of a post-education career.
Carnegie Mellon has offered a senior-level testing class for the last three years. Digital System Testing and Testable Design is taught by Dr. Shawn Blanton, an associate professor with strong interest in design for testing and microelectromechanical systems (MEMS) testing. We also offer a research program, and I teach some information about testing in my VLSI design class. So, our students are exposed to testing issues, but the focus is testing as a tool to producing very fine silicon, not simply testing in and of itself. - The testing industry must accept the notion that testing is part of the design. Testing is no longer a separate electronic domain. There needs to be more focus on how everything works together.
And I believe the right way is to position the whole problem as a vital, necessary component of education from the students’ point of view. Otherwise, it won’t work. If students don’t see the value, they won’t come to class.
About the Author
Wojciech P. Maly, Ph.D., is a Whitaker professor of electrical and computer engineering in the Electrical and Computer Engineering Department at Carnegie Mellon University. He holds a M.S. in electronic engineering from the Technical University of Warsaw, Poland, and a Ph.D. from the Institute of Applied Cybernetics at Warsaw’s Polish Academy of Sciences. Mr. Maly was elected a Fellow of the Institute of Electrical and Electronics Engineers in 1990 and has received numerous awards, and written many books, journals, and conference papers. Carnegie Mellon University, Electrical and Computer Engineering, Hamerschlag Hall 2126, 5000 Forbes Ave., Pittsburgh, PA 15213-3891, 412-268-6637, e-mail: [email protected]
Return to EE Home Page
Published by EE-Evaluation Engineering
All contents © 2003 Nelson Publishing Inc.
No reprint, distribution, or reuse in any medium is permitted
without the express written consent of the publisher.
February 2003