The tumultuous ride that pushed global economy down a path of recession starting in the second half of 2008 is not yet over, but is beginning to show signs of coming to an end. Arguably, the bumpy ride started a lot earlier than even that time frame. But when we do hit the bottom, which iSuppli Corp. predicts will be the middle of this year, the combined mass of the global economy will have shaken loose the dust that has been accumulating there since the 1920s when the Great Depression hit.
As the dust settles going into the second half of this year, iSuppli expects the market for semiconductors used in wireless systems will have contracted to industry levels between 2003 and 2004.
While all segments will face a contraction this year, the wireless area arguably will experience a worse share of the contraction. This area will be hit not only with soft end-market demand but also with inventory correction in the first half both, conspiring to yield a 26.7 percent year-over-year decline in 2009 revenues. However, for relatively healthy companies with scale, a down market provides an opportunity to identify and implement strategic alignment with upcoming market inflections and to further pressure competitors struggling to simply survive.
Achieving alignment is being accomplished through multiple means, including—but not limited to—strategic acquisitions/ partnerships for scale and expansion, repurposing of existing platforms for adjacent markets and continued R&D investments focused on next-generation technologies.
Gaining Momentum
One such inflection is the accelerating momentum behind mobile Internet devices and mobile consumer electronics. These devices and their requirements have created a veritable arms race in the mobile wireless semiconductor market. At the heart of this race is the fact that just as the devices fill the space between notebooks and smart-phones, their requirements also dictate optimized performance regarding processing capability and power consumption.
Where in their traditional target devices, chipset and system design had traded off these two parameters for each other, chipset solutions targeting this space can no longer afford to do that and still be competitive.
There are two main microprocessor architectures that are being used today and will continue to be used in the future by Mobile Internet Device (MID)-class devices: X86 and ARM. X86-based architectures traditionally have been used primarily in compute platforms such as desktops and notebooks coming from companies such as Intel and AMD. In contrast, ARM-based architectures such as those solutions from Texas Instruments, Qualcomm, STMicroelectronics and Infineon typically have been used in mobile phones.
Consequently, at least historically, X86 microprocessors are optimized for processing capabilities while ARM-based devices have been optimized for power consumption—in line with the requirements of their respective target markets. As one might expect, processing capabilities and power consumption are usually two parameters that are traded off for each other in the design process.
While not often talked about in marketing propaganda, this difference in historically optimized capabilities is the crux of one of the most competitive races ever to be contested by semiconductor chipset manufacturers. The issue simply is that MID-class devices are unique in that in most cases, they require the processing capabilities of compute platforms and the power consumption performance of mobile devices. This begs the question: Is it easier for a processing-optimized architecture to now solve the power consumption problem or for a power-optimized architecture to solve the processing problem?
To further analyze this situation, let’s consider the main drivers for solving each issue.
Quantum and DNA computing notwithstanding—with the former in its infancy and the latter still in research stage—and assuming high levels of architecture optimization in the original design, increasing processing capability is primarily driven by hardware architecture modifications such as process node advancements as smaller geometries are able to support higher concentrations of transistors in a single die, thus driving computational performance but not necessarily increasing power consumption.
These hardware modifications also can usually be advanced independently of system design of the host device. On the other hand, power consumption performance, while partially driven by process node advancement, is also dependent on often proprietary and patent controlled algorithms that take into account system level design and usage scenarios of the host device. Consequently, all things being equal, iSuppli believes that companies with a focus on the latter can achieve an advantage in more quickly being able to address MID-class device requirements.
Next-Gen Rising
Another such trend that iSuppli has identified as a key inflection point for which companies are preparing is the upcoming wave of upgrades toward next-generation technologies and networks. While these upgrades are not forecasted to really occur until 2011, product cycles require that in order to meet this window, semiconductor suppliers should now be aligning their R&D efforts to best catch this upcoming wave.
It is not just in the main baseband and applications processors architecture that this work is occurring. Innovations are competing now to best solve the nascent problems of these technologies in areas such as the transceiver, the power amplifier and the front-end design, not just in the mobile devices but also in infrastructure equipment such as the base station and the core transport backhaul.
Some key issues being solved in these areas are:
* RF design adaptability for dynamically changing RF environments.
* ase station power efficiency in highly linear applications.
* Design efficiency supporting backwards compatibility and multimode operation.
A correlation of forces has conspired to set the wireless semiconductor market back by almost six years during the last three to four quarters. But as we enter the second half of 2009, iSuppli expects a return to normal seasonality and growth into 2010 and beyond. With that return also comes opportunities to not only survive but to thrive for those companies properly positioned to take advantage of the trends, driving the next round of growth.
No comments:
Post a Comment