Today during Qualcomm[QCOM]‘s UPLINQ developer conference, Dr. Paul Jacobs, Qualcomm Chairman and CEO, answered some questions from the press after delivering his keynote address from earlier in the day. There were some interesting questions fielded. Feel free to watch the videos below if you have the time to listen to all of the questions. We filmed the Q&A session in its entirety.
The first question that was fielded pertained to the never ending subject of 28nm supply shortages and asked about any possible lost opportunities or leads that Qualcomm may have had prior to the shortage. Paul Jacobs answered by saying that 28nm is the newest generation SoC size, Qualcomm has a shortage because they’re on the leading edge of the tech, and the shortage was because demand was too high for the supply they could create, not enough machines to make them. They only expect to have a couple months of shortage and are continuing to work with TSMC to get more, furthermore multiple sources will be supplying chips by the end of the year. There is an opportunity for qc’s competition to take market share because Qualcomm can not fulfill the full demand and that there will be some impact, but that impact will be minor.
The second question asked whether or not Qualcomm should consider investing in their own Fab. Jacobs responds by saying they’ve considered it, and have enough money to do it, but want to retain the (fabless) model they have. This is because to create the fab and run it, you have to spend a lot of management cycles on it. Qualcomm doesn’t have to do that, freeing it up allows them to progress the technology instead. Qualcomm has the ability and talent to do it, but they prefer to stick to the horizontal fabless model, you can move faster, and it’s a better overall strategy. There’s a possibility, though, and they’re looking at different options to prevent future shortages. He wouldn’t rule it out in the future if it’s necessary, but it’s unlikely.
The third question asked Paul if and when Qualcomm’s architectures and chips would begin to show up in servers. Paul responded by saying that they are currently working on 64 bit architectures, but haven’t announced anything for servers, but there are opportunities there. Power and heat are huge in servers, so mobile processors make a lot of sense. No announcements yet, but Qualcomm is looking into it. servers are a clear opportunity, and it’s a green thing to do anyways, to lessen power usage of server farms.
The next question asked whether or not Qualcomm was working to address LTE incompatibilities between networks? Paul responded by saying that the issue is that there are so many bands for LTE, and that is because governments everywhere just started assigning bunches of bands that were available to use LTE in. Qualcomm is working on the front end to make it more agile, and improve how many bands a phone can support. A lot of the issues are up to the carriers themselves from a business standpoint. Qualcomm’s primary goal is to increase the number of bands a phone can support.
Another reporter mentioned that the battery problem had not been solved yet, and asked if Qualcomm has done any research in battery tech to improve life. Paul responded by saying that they look at all battery tech, most of what they do is to reduce power consumption of chipset and display, phones are getting better, but not fast enough. They also are looking at being able to recharge phone faster and more conveniently (wireless charging) frequently to top off the power. That’s the near term fix, and getting more power efficiency is the next thing going forward.
Next, a reporter posed a question about the future of Brew and feature phones. Dr. Jacobs responded by saying that they’re still a big market, but if the market for them shrinks, Brew will go away, but for now that is not the case, so Brew is still very relevant. There’s still a market opportunity for Brew.
Someone else followed by asking whether there was anything left to integrate on SOCs? Paul responded with a question asking how much of the phone can u fit onto the chip? Radio components, etc. Qualcomm is looking into it. The answer is probably more, not necessarily a class of things, but always looking at new radio tech and sensors, etc, so it is possible they’d be integrated at some point.
The next person asked if Qualcomm will make their own smartphone or tablet going off of the recent Google and Microsoft announcements. Paul responded after a slight pause by simply saying no. He followed that statement by saying that they had good partners and that they are good at what they do and that Qualcomm does their own thing well and they want to keep it that way. He admitted that Qualcomm hasn’t had a good history of creating direct products to consumers and selling them successfully and that Qualcomm has enjoyed more success as an enabler of technologies.
The final question asked if Qualcomm is trying to really promote the snapdragon brand, whereas before they didn’t really do marketing. Paul agreed that in the past they sold our chips to very sophisticated buyers that would invest in labs to test the chips (carriers, companies, etc). In addition to those channels, there are other channels (retail) to reach consumers. Consumers cant test the same way, so they need to market their performance (and it helps that real life tests prove their performance claims). Paul followed by saying that when he spends a dollar on engineering, he knows what he gets, but when he does it on marketing, he has no idea… and acknowledged that he needed to fix that.
Overall, the press conference was a pretty good one with some good questions as well as some other questions that repeated things that had been addressed during the last earnings call, namely 28nm issues. Those issues are mainly due to the fact that TSMC simply does not have enough 28nm capacity to fullfill clients’ demands. The perception that Qualcomm is having yield issues is ridiculous and completely unfounded. We thank Qualcomm and Paul Jacobs for the opportunity to ask questions and get pretty detailed answers. We look forward to UPLINQ day two.