Notice that Toshiba will continue manufacturing analog and CMOS image sensors internally, outsourcing to foundries only for bleeding edge digital and SoC.
This is the classic fab-lite strategy -- keep manufacturing the specialty stuff in-house, where internal manufacturing adds the most value, and go external for the commodity, bleeding-edge of Moore's Law stuff that very few companies can afford to keep investing in anyway.
Its a very different initiative taken by Toshiba Semi. Since it is moving towards specialize departments like Visual CMOS sensors and Analog Divisions, it might lead towards some more improvements for electronics industries.
Yes, fab-lite strategy seems to have become the norm rather than the exception. People were expecting the Japanese giants to be the last to embrace this fab-lite strategy. Fab-lite appears to be the way forward as the transistor dimensions shrink further. Do you guys think we will get to live in world where other than Intel, almost everyone else goes fab-less? Certainly, some of the niche players will need to operate their own fabs, yet, a landscape where the majority are fab-less appears likely.
With the fab-less trend becoming the norm, the foundries will gain somewhat stronger market power and bargaining power. We might get to see foundries controlling the timing and quality of certain products. We might get to see foundries uniting into one or more coalitions to obtain a stronger bargaining position against the fab-less design companies. Do you guys envisage a standardization process across different foundries on design rules check and such lay-out specs? I expect the foundries to do business via somewhat uniform protocols in the long-run. For an example, an IEEE standard on layout design for a given technology node?
This is becoming norm these days, to go fablite. TI went fablite couple of years back (atleast in Digital front). As the transistor dimensions shrink, and thus pushing the fab maintenance cost higher, many more companies will think on these lines.
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...