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 : Qualcomm Incorporated (QCOM) -- Ignore unavailable to you. Want to Upgrade?

To: quidditch who wrote (35437)	7/17/1999 2:38:00 AM
From: engineer	Read Replies (3) \| Respond to of 152472

DSPs....Ho Hum....TXN has been spouting DSPs to take over the world since 1983!!!!! TMS320 chips. So have they taken over the world yet, now almost 16 years later? Nope... Have they sold alot of them...Yup. does it make life much better for cellphones today? Perhaps..... Problem is that they have remained right on the edge of the technology curve where using them makes sense. the applications that they spout off about are all very high volume applicaitons and the power/size/cost curves are always right at the edge of what it takes to make it a viable product. they need an order of magnitude change in speed/power/cost in order to gain that market. the justification is still too large not to allow people like Qualcomm from making their own DSP core and working it down to a fine edge wihtout paying any margins to TXN. same story fot ATT, only not half as good as TXN. IMHO - DSPs are good for low volume high margin stuff right now. they need a fundamental change to get into very low cost consumer stuff. A good well done fixed layout DSP in a custom chip will always win out over them at todays power/speed/cost point.

To: quidditch who wrote (35437)	7/17/1999 4:49:00 AM
From: DOUG H	Respond to of 152472

>>>>repositioning of TXN to get out of PC chips and memory and into DSPs was quite the call. TXN made some announcements about six weeks ago about new generation DSPs working in conjunction with a new software platform the result of which, it was said, would accelerate new features (mmedia, data) into wireless. Thread tried to dope out at that time whether this constituted a threat; I believe that we tentatively concluded that it did not (because of nature of DSP as necessary complement to voice/data transmission via CDMA). But the tech heavyweights on the thread did not, as I recall, weigh in on that discussion.>>>>>>>>>>>>>>> steven, I'm no tech heavyweight but I do recall an article on TXN's acquistion of Libit. They produce DSPs for cable settop boxes. TXN wants to take advantage of the programable feature of the DSP as opposed to the ASIC which requires new production runs each time a feature is added. This allows the guys in the field to make adjustments to the DSP's via software as opposed to replacing an Application Specific Integrated Circuit. TXN's focus seemed to be on the cable business. Libit competes with BRCM in the STB arena. There was no mention of CDMA or wireless in the interview but I guess DSP's are flexible little creatures. Any corrections are welcome. D.H.

To: quidditch who wrote (35437)	7/17/1999 9:28:00 AM
From: Clarksterh	Respond to of 152472

JG#1, SM, MW et al - DSPs are nothing particularly special. They are microprocessors very similar to the microprocessor in your computer. The difference is that they are optimized, not for ordinary if-then-elses or straight addition, but for mathematical functions used frequently in signal processing (e.g. Fourier Transforms). Just like ordinary microprocessors, DSPs require software to work, and as fast as they might be on their kind of math, they are still pretty slow compared to hardwired signal processing. Their advantage is relative flexibility. Thus the CDMA chipsets contain a DSP in order to allow as many items as possible to be changed quickly and easily via software updates. However, the hardwired portion is still required because much of signal processing in CDMA requires too much speed to be done via software (in contrast, a much larger percentage of TDMA can be done via software). As time progresses, and the data rates stay the same, the DSP's will undoubtedly do more and more of the work on a CDMA chip set. As for is TXN a threat to the Q!: No, not yet and probably no time soon. Qualcomm has experience with all three things needed in a CDMA solution - the hardwired signal processing, the DSP, and the software which handles many of the protocols etc. TXN only has experience with the DSP, while the other parts, particularly the software, are the hardest anyway. TXN might have a better shot at an entry if it were possible to split the hardwired portion from the DSP, but Qualcomm is delivering systems on a chip (hardwired stuff and DSP on the same chip), so it looks like TXN is out in the cold for CDMA and probably will be for at least several years. Hope this helps. Clark PS 'ASIC' is just any chip with a very specialized purpose. Although DSP's are somewhat generalized devices by themselves, when incorporated onto one chip with hardwired CDMA processing they just become part of the ASIC. I.e. the MSM2300 or MSM3000 are ASICs which include DSP's for voice processing etc.