Physics Based Reliability Qualification
Joseph B. Bernstein
joey@umd.edu
1-301-405-0357
University of Maryland / Bar-Ilan University
The solid-state electronics industry is characterized by relentless pressure to expand and improve functionality, reduce costs and reduce design and development time. As a result, device feature sizes have shrunk to the nanometer range and design life cycles of most commercial products are less than five years. The industry trend is to have fewer designs comprise a greater percentage of the total parts sold. This leads to system designs comprising almost exclusively commercial off the shelf (COTS) parts rather than have custom ASICs. These parts are manufactured by the 'billion' so they have the tightest possible manufacturing controls and are therefore the 'best' possible quality that could be purchased for any application.
Until recently, semiconductor device lifetimes for most COTS parts could be measured in decades, which was essentially infinite with respect to the required service lives of most commercial applications. As the design rules shrink, power consumption increases and voltage margins become almost non-existent for the designed performance level. The lifetime and failure rate of these commercial parts are the ultimate casualty. Most large, high reliability systems are built with the assumption that electronic components will last for decades without failure. However, counter to this assumption, device reliability physics is becoming so well understood that manufacturing foundries are designing microcircuits for 3-7 years to fulfill the needs of most (>99%) of the electronic industrial needs. The Military, Aerospace, Medical, Telecom and other important industries will not be able to afford maintaining advanced systems based on current reliability qualification procedures.
This seminar will demonstrate how qualification for reliability from the manufacturers is not based on any physical known mechanisms and therefore overestimates the actual field reliability by factors of 10 or 100. We will show that through knowledge of the physics of failure mechanisms combined with specific qualification that can be performed by the manufacturer and the user can lead to more accurate reliability modeling. We will introduce a failure-rate based reliability simulation (FaRBS) approach that combines physical models with qualification data. This approach enables realistic prediction of the reliability of large complex systems in the field allowing for more accurate life-cycle costing and maintenance assessment.
JOSEPH B. BERNSTEIN
JOSEPH B. BERNSTEIN (SM) is a Professor of Reliability Engineering in the Mechanical Engineering department at University of Maryland, College Park and in the School of Engineering at Bar-Ilan University. He has appointments in Electrical Engineering and the Institute for Research in Electronics and Applied Physics. Dr. Bernstein received his PhD in Electrical Engineering from MIT in 1990 and he is actively involved in microelectronics device and systems reliability research and physics of failure including power device reliability, ultra-thin gate oxide integrity, radiation effects, MEMS and laser programmable metal interconnect. He supervises the laboratory for laser processing of microelectronic devices and is the head of the microelectronics device reliability program. Research areas include statistical interactions of multiple failure mechanisms in ULSI devices. He also works extensively with the semiconductor industry on projects relating to system qualification for reliability based on fundamental physics and circuit simulation techniques and on programmable devices and repair in microelectronic circuits and packaging. Professor Bernstein has been a Fulbright Senior Researcher/Lecturer and set up a joint center for reliable electronics at Tel Aviv University and Bar Ilan University. This center is a collaboration between industry, academics and government agencies from Israel and the USA.