Global Sources
EE Times-Asia
Stay in touch with EE Times Asia
EE Times-Asia > RF/Microwave

Qualcomm's chip chief looks to 4G

Posted: 01 Jul 2006 ?? ?Print Version ?Bookmark and Share

Keywords:Loring Wirbel? EE Times? Sanjay Jha? Qualcomm? cdma?

Jha: China has the capability of driving middle and high-end markets much more than people give it credit for.

Sanjay Jha is the classic example of the engineering executive who rises to a C-level post. Having joined Qualcomm Inc. as a design engineer in 1994, Jha is now CEO of the Qualcomm CDMA Technologies division (QCT), the group responsible for Qualcomm's IC and software sales. Qualcomm's mobile video and service technologies may get the media buzz, but QCT realized revenue of $3.2 billion in 2005, and for four years, it has been named the best financially managed fabless IC company by the Fabless Semiconductor Association. Jha sat down with EE Times to share QCT's plans for continued domination in 3G and emerging 4G wireless.

EE Times:Can you describe how QCT works within the Qualcomm corporate structure and how the arrangement benefits customers?
Sanjay Jha:QCT is a division within Qualcomm. We're responsible for all IC development within Qualcomm, and all software and protocol stacks that go with the chips. The software part is an increasingly larger business for us. More than 50 percent of our engineers are now working in software. My sense is that virtually everyone who is in the IC business now employs more people in software than in direct IC hardware because of the system nature of IC design.

Integration is getting to the place where you have to provide a complete solution. There's a division within Qualcomm corporate R&D where a bunch of the new physical-layer interfaces get developed, although some get developed within QCT-HSDPA [high-speed downlink packet access], for example. W-CDMA was also developed here. Combined, we and corporate R&D can make that a product and drive it to the marketplace. We have relationships with over 50 handset manufacturers.

Very often at Qualcomm, we take an end-to-end approach rather than a component approach. It's in that scheme of things that we in QCT fit in.

How much does it help Qualcomm to have licensees vs. watching an opportunistic development of chipsets that might look exactly like your own?
Everything that we do gets licensed out. We are not here to protect the IC business by using licensing as a barrier. CDMA would not have been successful had we not licensed it broadly. In W-CDMA, we have a vast number of licensees. Quite candidly, we've had a large number of competitors in CDMA2000remember that Texas Instruments competed with us; Intel acquired DSP Communications to compete with us; LSI Logic and VLSI Technology competed with us; Motorola used to have their own chipset; Nokia did until recently. We've always enabled competition.

They say that the handset market has turned into one for the big boys. In W-CDMA, there may be 20 licensees, but how many of them are real?
The Motorola Razr is now at a level of thinness and functionality that you cannot do unless you have lots of technologies in-house. You must have highly integrated baseband with multimedia integrated in those devices, highly integrated RF chipsets. You need to have Bluetooth, Wi-Fiso many technologies. It's just not possible to rely on one technology to be successful.

The cost of playing in this business is getting so highwe've not broken out how many engineers we have in this division, but I can tell you Qualcomm invests over $1 billion annually in R&D. If you're not prepared to invest those kinds of numbers in R&D, I don't know how you're going to sustain competitive advantage in the long term.

Looking at the debate over whether true mobile broadcast video is ready for prime time, do you ever feel you have to make controversial decisions about leading a market too far?
There is a clear recognition that you have to lead, but you can't lead by too much. The way we manage that is with strong carrier and customer relationships at all times. You can't just throw something at the wall to see what will stick. You must work with carriers as development happens and get their feedback. What we find, by and large, is that 80 percent of our notions are right, 20 percent are completely wrong, and we have to modify that minority to meet market demand.

The original model for GPS was as an adjunct function to the cellphone, and all of a sudden location service becomes a product in its own right. How do you predict what the mobile device will be for location services?
One thing to say is that the phone has become the dominant mobile platform. I always say voice remains the killer app. But what other types of functionality are we capable of absorbing in that device? Gaming could be a very important application. We support gamingwe have a partnership with ATI and use a lot of their graphics cores.

Will Asia lead in such markets?
You see leadership in areas such as the first color phones. There is a widely regarded view that Europe was leading in wireless applications. I always felt, even years ago, that Asia was leading. We've tracked each of these applications and asked how long it took for them to become popular in the United States. We found the United States is trailing Asia in many trends by two years. China, to me, has the capability of driving middle and high-end markets much more than people give it credit for. There's certainly a large low-end market, but the population is so large there that even the middle and high-end markets are interesting.

Do you try to work with the Chinese government in reaching accommodation over issues such as TD-CDMA?
China's standard right now is a little bit behind the development of CDMA2000 and the dominant 3G standards. This is not the first time the Chinese have taken this view. The objective is to have local, indigenous IPR. I see this as a positive move. You don't understand the value of IPR until you have your own.

The number of viable baseband players has been winnowed considerably, but it appears that some newcomers in the RF/IF market might be viable competitors for the RF portions of your designs. What do you expect to see?
We had more than 20 licensees for W-CDMA. In baseband, there may be a handful of competitors, but there is still room for surprises. TI and Intel are somewhat behind, and anyone with a good DSP core could be a player. In the RF chain, there is even more room for new ideas. The problem I see is the increasing R&D burden. You must get close to 8 percent to 12 percent of market share to be considered real in the handset market. Consolidation is guaranteed, but what worries me is how consolidation will occur, which is unclear right now.

Talk in the wireline world is that far too many OEMs have eliminated hardware design teams and turned all design work over to semiconductor partners or to ODMs. Is this a problem in either the infrastructure or handset world in wireless?
Most of the technology development has shrunk to the single-chip level, so it's not surprising to see the core of design moving to the semiconductor company. In base stations and infrastructure switches, we've seen many companies outsource not only their manufacturing, but the basic design of their trailing-edge products.

In the handset market, I am not as worried, since I see 50 manufacturers still innovating in a variety of ways. When Qualcomm designed its first handsets, no one else was building to CDMA. So in some senses, the handset market is healthier now.

What about MediaFLO and competing video-broadcast standards like DVB-H? There was talk at the Consumer Electronics Show that mobile video is overhyped.
Don't think in terms of Internet Protocol video clips and iPod with video. A consumer platform could receive up to 20 MediaFLO channels simultaneously, and the carrier networks will operate on a subscription model, like an MSO (cable-TV multisystem operator), not as a video-on-demand Internet Protocol archive clip service. That means innovation not only in real-time video processing, but in 3D graphics in the handset.

What about WiMAX and the various flavors of 802.16?
We are interested in general in OFDM applied to multiple markets, and we're looking carefully at mobile broadband access. As for WiMAX, 802.16d [fixed] seems interesting for backhaul applications in particular, although the economics of fixed broadband access still carry questions. For 802.16e [mobile], I am very circumspect. There is no spectral efficiency in the type of deployments being considered. It will require billions in new investment for effective mobile service.

Intel has had a heavy hand in promoting WiMAX, but we can look at the acquisition of DSP Communications and Intel's failure to convince anyone it is serious about baseband design. Does this make Intel a paper tiger, not only for 3G and 4G, but for WiMAX?
We have learned never to discount Intel. They have not always been successful, but they are a true competitor to Qualcomm. They have too much leverage and funding to not be taken seriously.

- Loring Wirbel
EE Times

Article Comments - Qualcomm's chip chief looks to 4G
*? You can enter [0] more charecters.
*Verify code:


Visit Asia Webinars to learn about the latest in technology and get practical design tips.

Back to Top