To: masa who wrote (6769 ) 8/4/2004 6:26:23 AM From: Maurice Winn Respond to of 12247 Masa, given the trends in cyberphones, QUALCOMM's market share of ASICs is going to be a LOT more than 25%. Cyberphones are all going to CDMA. Flarion's OFDM won't be taking over any time soon for the voice component in wide area networks though they seem to have a competitive product for data services and I don't see why they shouldn't be able to do voice, especially if they buy QUALCOMM's power control and other trick patents. GSM and GPRS are really not much use for data. GPRS filled a gap while the serious stuff was rolled out. GSM is good for text messaging, but not megabytes of data moving quickly and cheaply to cyberphones. GPRS isn't any good either because there isn't enough capacity in GSM networks. To get an accurate idea of QUALCOMM's market position, you should think in terms of 3G ASICs, not mobile phones and bear in mind that everyone is moving to 3G networks. I don't think QUALCOMM has much of a position at all in OFDM. Several years ago, I visited Auckland University and found there was a group there working on OFDM. My competitive threat antennae started twitching like crazy. We had discussion about this in SI. At an AGM we had somebody ask Andrew Viterbi about OFDM after the meeting [on a one to one basis]. He said that QUALCOMM had a few patents in the area, but seemed to me to be very offhand and non-committal and not wanting to say much, which struck me as odd. A year or maybe it was two later, he left QUALCOMM and took up with Lucent/Flarion which was sorting out OFDM. I suppose he had been keen on OFDM in QUALCOMM but for some reason, QUALCOMM wasn't so keen as he was and I suppose that means Irwin Jacobs. Having worked together for decades, it was quite a shift in the mobile world's centre of gravity to have Irwin and Andrew part company. That meant something. What it meant was that Flarion and OFDM were getting some serious backing and they seem to come up with the goods to enable mobile cyberspace in the data realm at least. Andrew Viterbi had also been very keen on HDR CDMA [high data rate] which became 1xEV-DO [meaning 1 x 1.25 MHz channel evolution, data optimized]. Being data optimized means more spectral efficiency and no latency problems; I think I have that right]. He was a strong proponent for separating voice and data and reading between the lines, I think he and Irwin had different opinions of the merits of mixed voice and data in the same channels. So, 6 years later, OFDM is starting to look as though it'll be a real product with actual customers from next year. It remains to be seen just how much more efficient OFDM is compared with CDMA. CDMA is hugely [orders of magnitude] more efficient than GSM, but even so, it has taken over a decade for CDMA to start to look like a serious competitor to GSM and it's only the advent of mobile cyberspace data and the demands on bandwidth and speed and cost which has propelled CDMA's universal adoption for that service. Since it was that difficult for dramatically better CDMA to overcome GSM, I don't think OFDM will very rapidly displace a great deal of CDMA. In the overall mobile phone business, QUALCOMM's chip share is Applish. But what's coming with 3G is where the fun is. QUALCOMM is dominant there and all mobile CDMA will involve significant royalties to QUALCOMM, so it's not really the same as Apple. This is indeed a place where you can expound on any wacky and wild ideas you might have. Feel free to lay it all on the line. Mqurice