Slide 28 is the Solution for European Carriers
I have been pondering the dramatic change between the analysts and shareholder meetings in February 2001 and the meeting last week with respect to Dr. Jacobs' treatment of WCDMA. My assumption is that two years ago Qualcomm was on the outs with the European carriers but now it is working well with them. The implications of the latter on Slide 28 are what could be very important.
Two years ago Dr. J devoted a substantial part of his presentation to a discussion of the Stated Reasons for Choosing WCDMA; he set up 5 supposed advantages of WCDMA and then proceeded to shred them – I have 4 pages of notes on his comments. Later in the meeting he stressed that you could do 120 voice calls simultaneously with 1x versus 90 with WCDMA in 10 MHz of spectrum (and 35 GSM calls); moreover, 1x and 1xEV-DO were three or four times more efficient in delivering data than WCDMA. After the meeting, I asked Dr. J why discounted cash flow models would not convince the European carriers that the predicted delays in WCDMA rollouts versus 1x would give 1x a major advantage. He seemed to agree that the models would show this and said that they hoped to work with the Europeans on this issue. I thought that adding a couple of zero revenue years to the WCDMA spreadsheets would hammer the DCF. We even heard rumors at the shareholder meeting that key people in Qualcomm did not think that WCDMA would ever actually deploy.
In contrast, last week Dr. J said WCDMA would “be a significant improvement over GSM/GPRS.” In the Q&A, he seemed to agree with Ben Garrett that more cell sites were needed for WCDMA but the first reason that he mentioned presumably would also be a problem for 1x – weaker propagation of the signal at 2.1 GHz. Even in discussing the need for more cell sites to provide data rates at 160 kbps or higher, he did so as if he was wearing kid gloves. Dr. J and Don Schrock seemed very upbeat about the reception that the Qualcomm-based WCDMA products would receive at Cannes this week.
What has caused this change in the Qualcomm attitude toward WCDMA?
A couple of years ago someone said that the CEO of one of the leading carriers had effectively told Qualcomm to give up on the idea of replacing WCDMA with 1x in Europe by saying: “The train has left the station; deal with it.”
A year or two ago, I had hoped that the Europeans might accept 1xEV-DO as an interim step to be used in the 3G spectrum until WCDMA was ready, hoping that they would later see that it offered all the data that they needed.
Rather than being bizarre, Ben, Slide 28 seems to be a solution for the European carriers. The train may have left the station but it only has a few cars and other trains with cheaper, more reliable proven technology are going to leave the station too.
To repeat, Slide 28 shows how GSM1x could work with WCDMA. WCDMA would be used in the major cities, using the 3G spectrum. In the less populated areas, GSM voice would run on 2G spectrum. In the areas in between, GSM1x would offer 1x/1xEV-DO in 2G spectrum. I note that I had to print Slide 28 to pick up all of the graphics. The following are my notes on Dr. J’s explanation of Slide 28 (I am using quotes for emphasis; although the words are close, they are probably not an exact transcription):
“We’ve been working with various operators looking at different ways of perhaps supporting the provisioning of third generation services but doing so at a lower cost and here’s one particular one – something we call GSM1x. When WCDMA is put out – because it’s at 2.1 GHz and it’s a new service – initially it will be probably an island – probably in the larger urban areas. … But then you have broad areas not covered. Well, one thing is you can fall back on second generation; if it’s a GSM area, you fall back on GSM/GPRS. But that’s a tremendous drop-down in the capabilities – particularly for data.”
“And so what we propose is that those other frequencies using the same antennas – cell sites – in fact fewer of the total, one can put on CDMA2000 with an expansion in voice capacity – considerable expansion – as well as the higher data rate capability because they can get by in less frequency than required for the WCDMA. It can be done in the existing frequency band and therefore you can get broad coverage across an entire country during the period before you are able to build out WCDMA across the entire country and reuse frequency bands and base stations that are there, getting a much better financial return. The GSM1x means that the CDMA2000 technology can be used with a GSM core network – the network centers that they already have in place can be reused for both the GSM side as well as the CDMA side. So a number are looking at this possibility – the regulatory issues, etc. We think it does provide a way of providing services across entire countries much earlier than could be economically justified if one only went with WCDMA. We’ll see how people react to that.”
In the Q&A, my notes indicate Dr. J said:
“I believe that what will happen is that there will be some fairly dense provisioning of cell sites in urban areas – a limited number of urban areas – to support higher data rates and then outside the question is what do you do. Well here you have the opportunity, again as one of the slides showed, of going to a 1xEV-DO type capability. Until operators begin to get experience with WCDMA – look at the experience with 1x and 1xEV-DO – they’re still not quite sure about how best to proceed ahead. I think looking ahead we have a very good opportunity. First of all, the sooner WCDMA gets out there.., we’re very pleased about that. But I suspect there has to be supplementing it – using lower frequency to get broader coverage – to make sure that these advantages are countrywide, rather just city wide.”
Ben is correct that EDGE would be a cheaper solution if it worked as advertised by Nokia et al. Why would the Europeans reject EDGE as a solution? It strikes me that one possibility is that there has been a major shift in the Scales of Credibility. When GPRS and WCDMA were being designed, Nokia, Ericsson and colleagues must have had tremendous credibility with the European carriers for having delivered GSM, which became the dominant world 2G standard. Since then, what has happened? Nokia et al. have delivered GPRS a year or two late, with data rates that I gather have been a great disappointment and have vastly underperformed what was probably promised. At the meeting, Dr. J mentioned that some had claimed 177 kbps for GPRS; I gather the reality is more like 10. Nokia et al. have been consistent, with similar delays and disappoint data rates (absent very dense cell site configurations) for WCDMA. EDGE appears to be more of the same.
Contrast that with Qualcomm’s superb performance, which should have moved its credibility way up in the eyes of the European carriers. 1x networks have been great successes in South Korea and Japan and are off to a good start in the U.S. 1xEV-DO networks are being launched. But in the eyes of some of the European carriers, I suspect that the key metric has been Qualcomm’s performance in WCDMA. At the meeting last week, Dr. J said that Qualcomm’s dual mode handsets have been tested broadly on WCDMA infrastructures in Europe and that the testing has gone well. Qualcomm seems to be the furthest along in providing dual mode WCDMA/GSM/GPRS chipsets and phones based upon the 6300 chipset will out at the end of the summer or in the early fall. The more powerful 6500 is not far behind. The European carriers have to be most impressed with this performance from a company that they know must have hated the whole concept of WCDMA. Don Schrock at the meeting last week seemed delighted regarding the upcoming Cannes demonstrations: “We have Samsung and LG and Sanyo who will be at our booth and have some very spectacular UMTS/WCDMA terminals, including GSM as well. So we think it’s going to be quite exciting for the Europeans to see the Qualcomm chipsets in these phones that are being brought to the market during the latter half of 2003.”
Back to Slide 28. It seems to be that what Dr. J has proposed allows one of more European carriers to take the lead in obtaining the necessary regulatory changes to permit a carrier to confine WCDMA at the outset to the large cities and to permit 1x and 1xEV-DO to be used elsewhere as Dr. J outlined in Slide 28. A lot of face saving could be involved for the Governments – they will get WCDMA in some areas and can hope that it will expand in the future as technological advances are made. But in the meantime, the carriers will be able to provide a cheaper more efficient service elsewhere based upon proven Qualcomm technology. The big losers would be Nokia and Ericsson. But one wonders why the English, Spanish, Portuguese or Italians would want to protect Finnish and Swedish companies from their mistakes. The Germans and French may feel differently if Siemens and Alcatel still have a lot at stake here. The carriers would probably welcome the prospect of cheap phones coming in from China as they gear up their 1x production.
All of this may not happen very soon. It may take a year of weak isolated and expensive WCDMA rollouts and more successful 1x and 1xEV-DO rollouts elsewhere before the regulators are willing to make the necessary changes. In many of its meetings and presentations in the past year, Qualcomm has mentioned that it is working closely with Vodafone. Vodafone has a major stake in J-Phone and is undoubtedly very familiar with KDDI’s turnaround and successes built on 1x. Vodafone has a major stake in Verizon and presumably has a lot of information about the 1xEV-DO trials in San Diego and Washington, D.C. Vodafone and Verizon own Iusacell which launched a 1x system in Mexico.
Last week, Dr. J answered a question about Vodafone’s interest in the 6300-based phones: “This allows them the opportunity to have a phone where they can have the GSM service in Europe and then people that roam in the U.S. can roam on the Verizon network, rather than on another GSM network here.” Don Schrock added, “Vodafone and Unicom are extremely excited about these chipsets. It can appeal to some high-end consumer that will roam from country to country. In addition, the RF chips that go in it will dramatically lower costs of GSM and CDMA in one phone – that’s one of the real strengths of the complete chipset. Then, as this takes off, what we are looking at for future generations is to segment that GSM/CDMA marketplace – not only the high end but, as you’re pointing out, come down to lower end chips also to address the markets throughout the world.”
It sounds as if Vodafone as seen a lot of Qualcomm. I hope Vodafone is one of the “number” looking at all that Slide 28 entails. |
