|
| Register for our Free Newsletters |
|
 |
|
|
|
|
|
|
|
|
| Other Carouselweb publications |
|
|
|
|
|
|
|
|
|
|
| |
| |
|
|
|
|
|
|
|
|
Seeking a new way forward
|
|
Component Obsolescence Group (COG)
: 19 June, 2003 (Company News) |
|
|
Professor Roulston, OBE, FRSE, FIEE, CEng is also Industrial Professor of Electronics at Edinburgh University. His work with BAE Systems is concentrated on the Avionics Group, and his views on the changing face of the electronics industry made a big impact when he presented the keynote speech that opened the recent COG conference on The Application of Obsolescence Strategies. |
|
Professor Roulston, OBE, FRSE, FIEE, CEng is also Industrial Professor of Electronics at Edinburgh University. His work with BAE Systems is concentrated on the Avionics Group, and his views on the changing face of the electronics industry made a big impact when he presented the keynote speech that opened the recent COG conference on The Application of Obsolescence Strategies.
Today's new age of electronics began, he said, with the so-called 'Perry Initiative' which, in 1994, proposed that military contractors should use performance-based specifications rather than MIL Specs, and use COTS (commercial off the shelf) products whenever feasible. The revolution that followed involved semiconductor manufacturers withdrawing from MIL Spec devices, commercial and professional components being used in MIL systems, the development of COTS embedding and open systems, the switch to plastic encapsulated Hi Rel devices, and a massive drop in the influence of the military markets. It was soon realised by industry that asking Intel to supply 300,000 microprocessors per year of the current vintage was equivalent to asking Boeing to supply one 727 every two years.
Today, the commercial market is pushing silicon towards ever-higher levels of integration and complexity. This is underlined by the recent announcements that 60nm LSI production is now on its way and, perhaps, it is time to ask whether there is room for another strategy? Over the past 30 years or so, we have seen five orders of magnitude growth in microcircuit complexity. But there has been a much smaller growth in system requirement!
Nanometer dimensions lead to new physics for semiconductors on silicon: Quantum and tunnel effects - chemical effects with time constants that matter - metal failure, oxide failure; radiation susceptibility, single event effects; thermal design that suits one chip per system, but becomes unwieldy for multi-chip systems.
Looking at COTS and trends, Roulston stated that, for a large high performance Hi Rel box, the COTS route is very much harder than anticipated. Reliability (ignoring SEE) is hard to maintain at past levels and there are risks carried forward that did not exist with more customised designs.
Although tools and techniques improve, the advantages are consumed by the extra complexity of making a product work in the specified environment and even more complexity, (e.g. error correction) may be needed just to stand still. We are vulnerable to uncontrolled feature shrinkage, which may invalidate an assumption that successive generations in a chip family can be facilitated through investment in design abstraction.
The Perry initiative has now run its course and we need a new way forward for the aerospace, military and similar professional equipment.
FPGAs, with appropriate tool sets, could become the backbone of military processing but the future is not going to be easy in the Hi Rel sector.
Q&As from the NOC Helpdesk Q What solution options are available to resolve a problem with an obsolete part? A A, There is a recognised sequence of resolutions based on a cost to implement basis and comprise of the following: existing stock, reclamation, grey market, alternate, substitute, after-market, life of type buy, emulation or redesign.
Q What areas should I consider as critical in a system? A
Materials with a platform history of obsolescence problems, predicted obsolescence problems, predicted wear-out, or industry acknowledged obsolescence problems such as microcircuits, software, bearings, fasteners or parts needed to be supported by National product qualifications.
Q Is uprating a solution to my obsolescence problems.? A
Typically, this is employed for temperature extremes and is one way of using commercial grade parts in non-benign environments. The downside of this option is that a detailed understanding of the original manufacturing processes is essential. Additionally, this practice will usually invalidate any assurances/guarantees offered by the original component manufacturer/supplier. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
