Lounging around in the Telecosm
I have met with Novell executives and am much impressed by the range of their Internet oriented storewidth technologies. I am also impressed by the aggressiveness and precision of their strategy, which is essentially to render most of the proprietary net enhancers from Akamai and others as open Internet standard programs on a Novell proxy server. I am convinced that directories are rapidly gaining importance on the net and that Novell commands the best most scalable directories. Whether they can accomplish their goals, however, remains imponderable and the stock is still a high risk high reward item, not to everyone's taste, but for turnaround seekers with strong stomachs it is very inviting. 5/24/00
If you are confident in the telecosm, the bears are for buying. 5/24/00
Sorry but I have been working on the new GTR and have not yet gotten around to answering T's insulting item, which continues the nearly ten year old TDMA tradition of ad hominem attacks on CDMA supporters (I am merely pumping a stock I own) and wild exaggerations of TDMA capabilities. I still have to check further, but on the surface the claim that ATT's TDMA and GSM naturally work together is spurious, since among many other differences one functions in 25 kilohertz channels and the other in 200 kilohertz channels. Moreover, as far as I know, EDGE resembles HDR in not being suitable for mixed voice and data communications, while CDMA2000 or WCDMA can mix the two datatypes. If TDMA could readily do that, why is ATT preparing to move to CDMA for 3G along with the rest of the world? 5/27/00
Qualcomm is already testing CDMA2000 chipsets, which are entirely pin compatible with the existing IS95 CDMA installed base. Qualcomm is a year away from producing WCDMA chipsets. Although WCDMA has arbitrarily incompatible specs, it is almost entirely based on Qualcomm IP, as the NEC experts acknowledged to me in Japan. It is quite possible that Qualcomm will prove to have the best WCDMA chipsets.
If the CDMA2000 system works as well as it should, however, it is possible that it will become the 3G standard. Regardless of the politics of standards bodies, in the end companies will have to use the system that works best.
5/28/00
We should do a chart at some point, but the differences reduce to the "direct sequence" form of the WCDMA coherent five megahertz band and the CDMA2000 three "direct sequences" of 1.25 megahertz bonded together. CMDA2000 comprises a 3.75 megahertz band to preserve compatibility with IS95 and HDR. Contrary to WCDMA claims, the Qualcomm 1.25 megahertz is nearly optimal and the wider spread of WCDMA still yields inferior performance to the bonded system. Unfortunately for the anti-Qualcomm forces, broadband wireless is still an "undershoot" technology (the market demands leading edge performance, not political claims).
If I could give some advice after eight years of immersion in these debates, I would urge a deaf ear to the endless detailed claims of superiority which the anti-Qualcomm forces will propagate. CDMA is hard, and Qualcomm engineers know how to do it. HDR shows they can even do TDM better than the GSM people can. (At Linkabit, the Qualcomm people invented TDMA also). Though HDR is code division between sectors, it is an adaptable dynamic TDM system for each connection. The reason for the TDM choice is the spreading code for a 2 megabit per second channel would be some 200 megahertz, which is not today technically feasible at reasonable cost.
In response to an earlier post holding that I lack a grasp of Gorilla theory and its stress on barriers to entry,
I admit I don't believe in legal barriers and consensual standards. Execution is absolutely vital; standards, like history texts, are written by the winners. 3Com, Intel, Microsoft, Applied Materials et al won not because of some legal monopoly but because they moved first and sustained their learning curve ahead of all followers. 3Com failed not because the Ethernet standard failed but because Synoptics outperformed them with the innovation of 10BaseT (Ethernet over twisted pair). Qualcomm will win because they are masters of the technology that all agree will be the foundation of the wireless internet. WCDMA is not a significant innovation; it is a political play. In the end, as I said, politics will give way to the exacting practicalities of creating broadband CDMA systems that can handle voice and data robustly at once.
5/29/00
Yes, Greenspan's perverted yield mountain could bring about a recession and yes, faith in the telecosm companies can move mountains. The optics, wireless 3G, and storewidth paradigms are still embryonic.
5/29/00
Ovonics has been all smoke and no turkey for the last twenty years, but Tyler Lowry is perhaps the world's leading expert on semiconductor memory technology. This is a company to watch.
5/31/00
Beta was actually inferior in the crucial dimension: ability to play two hour movies or sports events. Gratuitous superiority gets you nowhere. You must be superior in the features that the customer values. GSM was superior when it was launched. WCDMA may well prove to supply less capacity, data, and robustness than CDMA2000, which Qualcomm is constantly improving. If so entrepreneurs will use it to compete with WCDMA.
5/31/00
Gee, whiz. After eight years of being described as the vendor of a Qualcomm securities scam in violation of the laws of physics, now I am extreme and hyperbolic for pointing out the flaws in TDMA. The next post describes the rollout of CDMA2000 in Korea at the end of this year. TDMA EDGE is at least two years behind and five times slower. It will not happen.
My critic implies that the Qualcomm advantage is small. It is large and growing larger. Perhaps the Japanese will manage to concoct a competitive version of WCDMA before Qualcomm does. We will see. But the power efficiency, capacity, and bursty data capabilities of the CDMA side will become increasingly evident in both mobile and wireless local loop applications.
LEAP wireless has interesting plans to demonstrate the superiority of CDMA in wireless local loop. In Chatanooga, the company has already taken over 7 percent of the fixed local loop market with a $30 per month all you can use product that also can be employed as a mobile. The local loop is where the capacity limitations of the ATT system will be most crippling.
GSM does have huge volumes and will be an important force. But GSM is not going to pursue EDGE in the face of a Qualcomm demonstration of a fivefold more powerful system, available earlier. As for AWE, it just has a huge marketing budget and alot of ad hominem arguments. It will not prevail. 6/7/00
********************
Alot of defensive writing for CDMA. Haven't had much success in using the site during the Memorial day week. Working well today though.
Jack |
