Monthly Archives: November 2012

Starting  with a low cost/low complexity  design alternatives  in large engineering programs  and then  systematically raising the cost and complexity , where warranted,  is a far better approach than starting with higher cost design complex designs  and later doing cost down and simplification  activities.  I have worked on many product teams where the latter was the norm.  Subsystem teams would select costly more complex technologies, more costly high precision components and assemblies, costly manufacturing and control systems, all  in an effort to get  a jump start on functional performance and time to market requirements. Early demonstration of performance, even though the costs were  over allocations, were considered perfectly fine,  as long as everyone tacitly understood that  there would be a cost-down and simplify  activities  at the end of the development activities.  Many engineering teams  understood  that their cost allocations would be waived initially to satisfy time to market and performance requirements.

The higher initial cost design approach was quite appealing,  as it usually  avoided  the unwanted attention and pressures from engineering management.   Company buyers, in turn, would prepare themselves to put undue amounts  of pressure on suppliers to reduce costs for their deliverables.  Manufacturing engineering would invest time and resources in higher precision and many   processes with secondary operations, again with the idea that both design and manufacturing cost down would come later.     Sales and marketing people, in turn,  would prepare for  a higher priced offering than originally planned.  Extra pressure was put on sales teams to push the higher prices along to loyal customers.

The high cost design approach many time proved difficult in later development stages  as the cost down would negatively affected performance. This is really something one  would like to deal with early on in the design cycle, not at the end.   Rationale for original design choices were sometimes lost or forgotten.  The cost engineers  tasked with the cost down and simplification activities were usually not the original design engineers.   This in turn created new difficulties usually for downstream service engineers and manufacturing quality engineers.

When starting with low cost design alternatives, it becomes imperative  to quickly identify a set of robust technologies and robust manufacturing processes that simultaneously satisfy quality, cost, and delivery requirements.  In selecting low cost alternatives, engineers are tasked with   exploring  available design space (using flexible fixtures)  to identify first a working  prototype condition and  directions for  improvement without adding cost. Using experimental design methods /parameter design methods  to find an optimal set of nominal values, has been widely used.  The rule of thumb was that if you could get the functions to work just once with the low cost  approach, then you could begin the optimization process without adding cost.  There would be many opportunities to capitalize on  better combinations of control factors and signal factors.  If the optimization efforts fell short,  then adding cost incrementally until the trajectory to design maturity  improved, could be done.  Nevertheless, the initial low cost approach would still end at a better place than starting with high cost and trying to drive the cost and complexity down late in the cycle.

When we introduce a new chip, we plan and execute a comprehensive reliability qualification plan. This plan will be based on many different reliability stresses addressing infant mortality rate test, early life failure rate test, long term life test failure rate prediction based on a small population of samples pulled from early production lots .
Due to the fact of limited device sample sizes, we are trying to assign a confidence level to our failure rate predictions using “industry standard” chi-square adjustment in the hope, our prediction will be closer to real field failure rates.
This is a “standard” approach of the semiconductor industry because testing very large sample sizes of chips is economically not feasible, especially for small-and fabless semiconductor companies.
IBM Corp.’s Semiconductor Division calls the above practice “finding the tip of the iceberg  “only indicating if there are major catastrophic failure mechanisms”. IBM and major semiconductor manufacturers are stressing large sample sizes in ongoing reliability testing of the outgoing device population.
Above approach requires capabilities and facilities for ORT (ongoing reliabiliy testing) of tens of thousands of devices per year. Only major dedicated manufacturers do this
(like Intel, National Semiconductor, Micron, etc. )
In the course of 2-3 years of intensive ongoing reliability testing of samples of the outgoing population combined with field failure information will one be able to make reliability assessment and meaningful prediction of the maturing semiconductor product.