«

»

Sep 19

QRD:Qualcomm’s Secret Weapon (Edition 2)

Brief Introduction

Purpose: As the coming of commoditization of smartphone is inevitable, Qualcomm needs to prevent the replay of MediaTek in smartphone. Once MediaTek dominates the middle-to-low-end smartphone, it can attack high-end market. Qualcomm develops QRD(Qualcomm Reference Design), a turnkey total solution, and stores it in warehouse, being ready to give low-cost chipset vendors such as MediaTek a deadly blow.
Progress: This project has been going on for several years in low key. One hundred Chinese engineers have been put on this project in Shanghai. However, so far the project does not run very successfully and has failed on multiple platforms such as 6240, 6270, 7225 and 7227. One of the reasons is that the previous software platform was BMP and ran on feature phones. Another reason rests on Qualcomm’s requirement that stability and quality was the most important, which led to high cost. Now Qualcomm is launching new platform 7227A. In contrast to before, Qualcomm now tries to seek balance between stability and low cost and also to shorten time span between non-QRD chipset and QRD chipset.
Final Pyramid Infrastructure: Top layer is occupied by Qualcomm who is responsible for pushing the whole project, leading in reference design, and providing technical and capital support. The second layer is occupied by design house. Unlike design house in era of featurephone, design house for QRD can be categorized into two types, hardware design house and software design house. The former concentrates on the design of PCB while the latter focuses on software platform. The third layer is taken by third-party application integrators who integrate diversified internet applications and offer them to consumers. The fourth layer is taken by thousands of SME of handset making who are expected to leverage their low cost and rapid response to market demand to boost the huge growth of smartphones in developing countries.

   Opportunity
Qualcomm:
Ideally, Qualcomm will be one of two dominants of industry chain once QRD makes success. If MediaTek intends to grab a potential client, Qualcomm will show the client the same or even better solution with the same price. The relationship of Qualcomm and MediaTek may be much like that of Intel and Via.
Software Design House: Thanks to the severe shortage of human resources in the field of Linux, especially Linux Kernel for portable devices, software design houses will probably be the second largest beneficiary from QRD and maintain their competitive advantages of close relationship with Qualcomm, scalability, wide adoption, and rich experience. What they ought to do now is to take actions to get as many Linux talents as possible as a preparation for the boom of low-cost Android smartphone. These actions include recruiting present Linux talents and fostering new talents in training schools. Pay attention to these companies which deserve investments.

   Problems To be Solved
Revision of Qualcomm’s Accounting Methodology:
So far Qualcomm has not successfully established a benefit-sharing ecosystem. Qualcomm is employing a strategy of purchasing rights of partners’ products and building a total solution on its own rather than cooperatively building a total solution and sharing profits. One of the barriers is the inflexibility of Qualcomm’s internal accounting methodology which has never allowed the flow of partial profits of chipset to other companies. Other players will not heartfully support QRD until they are convinced of their share of benefits from sales of total solution based on Qualcomm’s chipset.
Degree of Qualcomm’s Attention: QRD team is one department of QCT which cares about the sales of all chipsets rather than only QRD chipsets. Hence, the revenue of QRD is not the main criteria to measure the performance. Instead, the presence and market share of low-cost chipset players such as MediaTek is the worry of the team. 6573, a 3.5G chipset, which is planned to launch in June by MediaTek, has been postponed due to a big problem on modem side. But the performance of AP in 6573 is impressive. The contribution and value of defensive weapon and people is usually underestimated. Qualcomm should put more attention on QRD.
Removal of Cultural Gap: The success of QRD will be highly dependent of Qualcomm’s cooperation with Chinese local companies. Owing to huge gap between western culture and Chinese culture, an excellent leader of QRD team who know how to persuade Qualcomm’s senior managers as well as deal with Chinese local vendors is vital.

    Q&A
Q1: The average cost of labor of Qualcomm is much higher than that of MediaTek. Is QRD feasible?
A1: Although labor cost of Qualcomm is high, Qualcomm does not provide turnkey solution only by itself. Instead, it chooses to cooperate with Chinese software design houses which can leverage the cost of Chinese engineers to make customization.
Q2: Will QRD pull down the price of smartphone very soon?
A2: No. Qualcomm won’t employ a strategy of “attack” by QRD because it will harm Qualcomm’s profits. Qualcomm regards QRD as a defensive weapon to beat low-cost chipset vendors.
Q3: Would frequent update of Android version be a threat to QRD?
A3: Frequent update of chipset and operating system can be a threat to QRD, bringing about an incremental cost of redesign and debugging. However, as chipset and OS become more and more mature, update will slow down and hence the cost will accordingly decline.
Q4: The entry fee for handset maker is high. Can it fall in future?
A4: Of course. The entry fee is usually high at the beginning but will fall rapidly if more and more participants recognize the value.
Q5: Will MediaTek vertically integrate design house and respond more quickly and cheaply than Qualcomm?
A5: Theoretically, MediaTek can leverage its advantage in culture and management to integrate Chinese design house as it ever did in Shanzhaiji. However, unlike the enclosed ecosystem of featurephone, the open ecosystem of smartphone needs professional division of labor and massive innovation. Vertical integration may not be a good solution, which can be observed from the history of PC.