Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : IDTI - an IC Play on Growth Markets -- Ignore unavailable to you. Want to Upgrade?

To: Steve Lewis who wrote (6364)	2/10/1998 1:24:00 PM
From: Rob S.	Respond to of 11555

Thanks for the recent posts. Brian B. had said that they have been considering foundry relationships including one with IBM. With NSM now owning Cyrix, that leaves a vacancy at IBM for IDTI to enter. Cyrix's huge, complex, 5 metal layer M2 never achieved high yields in the IBM plants and has lost money. The simpler, smaller C6 will achieve much higher yields. However, IBM will extract a level of profit from anything they produce for IDT/Centaur. That does not matter nearly as much as being able to have a second source for the parts. As you pointed out, major OEMs will be much more comfortable buying the C6 if they know that it is alternatively sourced. Intel has their own "second source" capability by virtue of the fact of their long tract record and multiple production facilities. A couple of additional benefits would come out of a relationship with IBM. The first one is that IBM is fully cross licensed with Intel and would shield IDT from a potential lawsuit from Intel; just as it did for Cyrix (Intel sued but Cyrix won). Nothing has been printed that I'm aware of, but I suspect that IBM also has the ability to produce slot 1/slot 2 Pentium II type architecture under the cross license, similar to the arrangement with NSM. That would allow IDTI to enter the market for Pentium II compatible modules. The second potential major benefit is that IBM is the leader in copper metalization technology. This is starting to look like it will be an advantage in next generation uPs for two reasons: 1] it offers better conductivity and lower parasitic reactances; You can lay in finer line interconnects and design the parts for higher frequencies. IBM recently announced development of 1 GHz RISK uPs and that they expect to be in production around the year 2000. The size of the chip is reduced, the power disssapation is reduced, it is better suited for parts with on-chip DRAM with fine granularity. The C6 core is optimized more for high frequency (witness the 3x clock) than it is for complex, (some would say convoluted), architecture. The idea is that it can be more easily run at higher frequencies - although we haven't seen results of that yet. My guess is that it could be adopted well to copper metalization - better than its competitors. Higher frequencies are meaningless unless you have higher frequency I/O capability either through going toward on-chip memory and systems-on-a-chip (SOC) and/or going toward a multi-chip module architecture like the Intel slot 1/2 scheme. Slot 1 is only a partial solution and is already running out of overhead. SOC is the future for high volume deigns. Doing a deal with IBM comes with a hefty price tag. IDT will make more incremental profit from producing parts in their own plants. But IDT must consider the overal strategic picture - being your own best competition via cooperative fab and development arrangements is crucial to success. Perham and the board should keep in mind that their is a bias built into their operations for perpetuating the "roll your own" mentality. Key managers' first gut reaction to cooperative ventures with outside parties may be that it dilutes their own authority and prospects. At the extreme that thinking creates a scenario in which the steerage of the ship is not given up even though the ship is headed onto a reef. As I've suggested before, IDTI should consider having a combination of internal and external fab capacity and should seek strategic alliances whenever it makes technical or marketing sense combined with even minimal financial sense. I think a deal with IBM could be great. The street would warmly welcome it. Caching, caching (show me the money!).