Nuts & Bolts of Semiconductor Lifecycle Management, Part 1

In my most recent column I pointed to the financial business values of an entirely new model for the military/aero component supply chain with the acronym SLiM, for semiconductor lifecycle management. We will look at the new model in greater detail over the next few columns. (See: Requirements for a Successful IC End-of-Life Planning.)

With aerospace system lifecycles ranging from 20 years to over 50 years, IC volumes clearly cannot drive the semiconductor market. Then factor in Moore’s Law and quickly evolving consumer electronics markets with 18-month product cycles. Commercial off-the-shelf (COTS) and International Traffic in Arms Regulations (ITAR) initiatives leave military/aero contractors more exposed. Further challenges include semiconductor suppliers exiting the aerospace and defense market. It’s a pretty gloomy picture, but it gets even worse!

Market dynamics force aerospace and defense contractors to get involved early in end-of-life management. There are specialty resellers and distributors offering bill-of-materials management programs. However, opportunistic dealers and traders offer end-of-life inventory at highly inflated prices. On top of this, semiconductor counterfeiting is an increasing problem. There is clearly a need for a better way. A perspective on obsolescence mitigation (OM) was laid out in earlier columns. OM is clearly not a solution to the challenges of consistent, reliable supply. (See: Get Strategic With Military/Aero IC Supply Challenges and Military-Aviation Semiconductor Sourcing Remains Problematic.)

There are three distinct stages to SLiM. Stage 1 is the early design phase for high-reliability products being used in military/aero applications, when it is essential to establishing the long-term supply chain. Stage 2 involves wafer banking, and Stage 3 applies to a foundry transfer and/or product reengineering strategy. See the following graphic for how it occurs over time.

Three Steps to SLiM

Stage 1 has three phases to settling the supply chain. The first phase involves partnerships with major commercial semiconductor manufacturers, as I advocated in my last column. In this initial phase, licensing agreements are created allowing military/aero-focused suppliers to serve aerospace and defense markets. Some of these agreements already exist. The agreements include early access by commercial suppliers to their roadmaps and products and also can include joint marketing activities.

The second is the engineering phase, which involves creation and qualification of high-reliability versions of commercial devices including electrical characterization in high-temp, ruggedized packaging.

The final production phase involves manufacturing and qualification to QML Class Q and V requirements. This requires high-performance testers, high-reliability screening, burn-in, and life testing, making the devices truly “flight ready.”

In my next column I will address Stages 2 and 3 details of semiconductor lifecycle management. Wafer banking, foundry transfer, and redesign strategies that can securely, reliably, and cost-effectively extend semiconductor device lifetimes beyond 20 years will be discussed.

Semiconductor lifecycle management is a multi-decade, robust supply chain approach. It requires early and well designed partnerships with key commercial semiconductor manufacturers. Also essential is a razor-sharp focus on the aerospace and defense electronics market. For success, a three-stage supply chain approach, as outlined above, should be adopted. Ideally, suppliers wanting to play in this arena should also have an operational presence in both Europe and North America, the two epicenters of military/aero electronic systems design.

5 comments on “Nuts & Bolts of Semiconductor Lifecycle Management, Part 1

  1. Barbara Jorgensen
    February 9, 2011

    You make a great point about the difficult relationships between commerical suppliers and mil/aero suppliers. In many cases, their end goals are entirely different–serving the quick-turn consumer market vs. serving the long-term mil/aero/medical markets. All markets are risky, but mil/aero has the added issue of being dependent on governments that in the case of the US could change every four years. Yet, a design-win in the mil/aero market can be a cash cow for decades. It's a really complex situation. Thanks for offering some advice and some actual how-to steps to help guide suppliers and customers through the concepts.

  2. elctrnx_lyf
    February 14, 2011

    The thee stage approach to handle the challengs of obsolscence mangament definitely makes lot of sense. The first stage is definitely a crucila step to identify the right suppliers with the good amount experience in making the parts for aero/defense market. If there is more than 100 suppliers and 100 companies making the aero/defense products are pool themselves up and can come out with a right strategy for the future and can deliver the high quality products.

  3. prabhakar_deosthali
    February 16, 2011

    In the comments to the blog

    Requirements for a Successful IC End-of-Life Planning.

    I had suggested an alternative to Wafer banking, that of making the designs portable. By provisioning for the changes in the components at the design stage itself , the problem of end-of-life management can be somewhat tamed.  At the design time , the product designer should lokk at the component road maps of the component companies . This will help the designer in making suitable design considerations to accomodate the newer components ( with may be new form factor, new material etc) in the existing design.  20 years is offcourse a very loooong period considering todays semiconductor evolution pace. Becasue in such time span many a technologies may become extinct. Even if you go for wafer banking, the associated  packaging and testing tools may not remain in existance! The defense and avionic industry has to live with this hazard.

  4. J-TX
    March 3, 2011

    That's why the Mil / AERO OEM needs to develop a relationship with a good supplier that specializes in Diminishing Source Management and Mitigation.  For example, at Falcon Electronics, I often engage with customers looking to either source product that has gone obsolete, or mitigate the effect of an EOL.  Luckily I have a franchised partner, DPACI, which does die banking, re-packaging and test, even die harvesting and re-packaging.  We have 2 partners who test and screen brokered parts at multiple levels to detect counterfeits.  Yes, it's complex, but the resources to make it work are out there, and they don't always break the bank.

    I encourage those with questions to contact their Falcon Electronics Rep.

  5. ddeisz
    April 5, 2011

    Until brand new development contracts are required to have an authorized long-term (20+ years) solution in place up front, the myriad of solutions being proposed will all be after the fact. While a long-term solution is usually part of an initial award, it still isn't as funded and a  required entity with enough emphasis. It just hasn't been elevated enough in most cases. Once that is happening on a regular basis, many of these proposed solutions will be a natural part of development.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.