Curiously enough, cost considerations have proven to have little effect on the outcome of standards battles in the past. In telecom as well as in storage, computer interfaces, and consumer electronics industries, standards with an early cost advantage disappeared and faded away, while pricier technologies got the upper hand. Cost efficient technologies can lose ground to more expensive alternatives for various reasons before lack of scale levels down their early cost advantage.

This cycle usually leads to the extinction of defeated standards. Examples are numerous: BetaMAX, SCSIs for interfacing computers components, and Hiperlan for Local Area Wireles networks, to name a few.

Although wireless is a market rich in niches, the possible survival of a defeated standard could be very much defined by its cost. After the successive salves of operators’ announcements about their selection of LTE, the question of the survival of WiMAX is completely relevant. Many in the industry are now thinking that WiMAX may eventually disappear altogether, or else remain restricted to a small niche.

With the significant momentum that LTE is gaining, the battle could be well over if WiMAX can't retain the cost advantage despite its lack of scale. WiMAX market dynamics could help overpass the scale limitation and help bail the standard out. WiMAX’s ability to compete in terms of cost may define its future and eventually ensure that WiMAX could still be an option for service operators in some emerging markets, for vertical applications and other niches.

Two main factors have proven to significantly impact the cost of a technology: IPR costs and scale. Obviously, the intellectual property licensing scheme has the potential to alter the cost equation.  Likewise, scale lowers the implementation costs of a given technology and initiates a virtuous cycle. The larger shipments of a certain technology, the lower the costs and the greater the potential to dominate the competition.

Despite the momentum gained by LTE today, the deal is not sealed –the cost equation is still very blurry. There is no indication of an IPR cost advantage for LTE. Similarly, the scale of existing WiMAX deployments has already allowed some significant cost reduction for WiMAX CPEs and base stations. Price erosion has been fairly strong so far in the WiMAX space, compared to other technologies.

Device prices are the most important factor, and they are usually driven by chipset prices. Last year, full WiMAX chipset modem (including baseband and RFICs) prices were as low as US$23 per unit for high volumes of the mobile MIMO type. In 2009, many expect mobile WiMAX silicon shipments to number over 3 million units. It is likely that prices could decline to under the US$20 point for high volume POs.

I believe this level of price erosion for such low volumes is unique. Much of this could be explained by the incredible dynamism of WIMAX chipset companies and the bubbling competition amongst them. It has also to do with the structure of the WiMAX device market, where Taiwanese and Korean ODMs play a determinant role. Many of those who built WiMAX devices were either new entrants or companies who were able to achieve a cost leadership position in the WiFi-DSL gateway market, like Gemtek and Zyxel, in cable setup boxes like Motorola, or in the wireless market such as Huawei and ZTE.

According the WiFi alliance, WiFi chipset prices reached the US$20 point when shipment got close to 50 million units.

According to iSuppli, the Infineon HSDPA modem costs Apple just over US$15 for its sensational smartphone, and it is selling at a rate close to 6 million units annually. Many of the 180 million smartphones expected to be shipped in 2009 will have HSDPA modems. In 2010, 18% out of total 1.5 billion handsets shipped will be HSDPA enabled. Today an HSDPA chipset solution ASP costs between US$15 and US$20. This is the price level that we expect WiMAX chipsets to hit in 2010, for shipments well below 10 million units annually.

It is likely that LTE will more closely follow the track of HSDPA in the ASP/volume curve than WIMAX. The LTE chipset market is likely to be structurally similar to HSDPA's. Main buyers and suppliers are likely to be the same in the smartphone space, as well as in the add-on cards segments. Vendors such as Qualcomm, ST-Ericsson, Texas Instruments, Infineon, and Broadcom are likely to dominate on the supplier side. Nokia, RIM, Apple, followed by Samsung, LG and Sony Ericsson will likely be the dominant buyers in the smartphone sub-segment. Huawei, ZTE, Sierra Wireless, Option, Novatel and to a lesser degree Taiwanese ODMs such as Foxconn, Quanta, Asus, Zyxel, and Gemtek will dominate the LTE add-on cards sub-segment.

While some of the players listed above also have a stake in the WiMAX game, the potential overall LTE ecosystem has more similarities with HSDPA than WiMAX, despite the early focus of operators on laptop users.

Although this is highly speculative, one could still reasonably expect LTE chipset and device prices to be higher than WiMAX's for a given shipment volume. The difference between the value chain components of WiMAX and LTE is likely to offset any LTE scale advantage from a cost standpoint.

WiMAX is truly at a turning point today. Its time to market advantage is being offset by the incredible dynamism of the LTE community and the scale that it has achieved in record time. Cost advantage is the last piece holding the machine together. Pushing down costs and margins could be the only way out for pure WiMAX players, whether chipset, device and infrastructure vendors, or WiMAX operators, to ensure they reap some fruits from their 4-5 years of intensive work and significant investments. Spirits may be low, but everyone should keep in mind that as WiMAX subscribers ramp up over the next few months, much distance could still be covered in the volume/price curve. Having enough drive and courage to go further in pushing volume up and price down could be life saving.

For more information you can contact the author at bilel@maravedis-bwa.com

Copyright © 2009 by Maravedis Inc. All Rights Reserved.
No reproduction without consent.



>> Top