One-two punch knocks wireless semicon market back to 2003

EL SEGUNDO, USA: The combination of the global recession and excess inventory has set back the wireless semiconductor market to levels not seen since 2003, according to iSuppli Corp.

Global revenue for wireless semiconductors, the majority of which are used in mobile handsets, is set to decline to $39.4 billion in 2009, down a stunning 26.7 percent from $53.8 billion in 2008. This is the lowest level of annual revenue for the wireless semiconductor industry since it generated $34.7 billion in 2003.

“While all major application markets for the semiconductor industry will contract this year, the wireless area arguably will post the worst performance,” said Francis Sideco, senior analyst for wireless communications at iSuppli. “Wireless semiconductors were hit in the first half by the one-two punch of soft end-market demand and an inventory correction.”

The figure presents iSuppli’s forecast of global annual wireless semiconductor revenue. This figure consists of revenue from sales of all semiconductors used for wireless applications, including mobile handsets, wireless infrastructure equipment, WLANs and connectivity products.Source: iSuppli, July 2009

While 2009 will be a poor year by any measure, conditions are improving markedly in the second half.

“As the wireless semiconductor market enters the second half of 2009, iSuppli expects a return to normal seasonality and growth, which will continue into 2010 and beyond,” Sideco said. “With that return comes opportunities to not only survive but also to thrive for those companies properly positioned to take advantage of new product trends, driving the next round of growth.”

MIDs deliver max opportunity
One new product trend is the accelerating momentum behind Mobile Internet Devices (MIDs) 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 MIDs fill the space between notebooks and smart phones, their requirements will also dictate that performance be optimized for processing capability and power consumption.

However, although these two parameters were traded off by chipset and system design in traditional devices, chipset solutions targeting this space can no longer afford to do the same sort of exchange and remain competitive.

Two main microprocessor architectures are being used today and will continue to be employed by MID-class devices: X86 and ARM. X86-based architectures from companies such as Intel and AMD traditionally have been used in compute platforms such as desktops and notebooks.

In contrast, ARM-based architectures such as those solutions from Texas Instruments, Qualcomm, ST-Ericsson and Infineon typically have been used in mobile phones.

Consequently—or at least historically—X86 microprocessors are optimized for processing capabilities while ARM-based devices have been tailored for low power consumption—in line with the requirements of their respective target markets.

ARM vs. X86
“The difference in historically optimized capabilities between the two architectures represents the crux of one of the most competitive races ever to be contested by semiconductor chipset manufacturers,” Sideco said.

“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?”

iSuppli believes that companies with a focus on optimizing the balance between these two parameters for specific target devices can achieve an advantage by more quickly addressing the requirements of MID-class devices.

Next-gen rising
Another such trend that iSuppli has identified as a key inflection point is the upcoming wave of upgrades for next-generation technologies and networks, such as HSPA+ and LTE, as well as the corresponding backhaul and upgrades of core networks that are required to fully take advantage of the advances in these radio access technologies.

While these upgrades are not forecasted to really occur until 2010-2011, 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.