By Mark LaPedus
The next-generation 4G wireless standard known as long-term evolution (LTE) presents some new and difficult design choices for OEMs.

One of the more difficult choices involves the less glamorous, but arguably the most critical part in a handset—the radio-frequency (RF) front-end. Typically, the RF front-end often comes in a module and includes various key components, such as the power amplifier (PA), antenna switch and filter.

The latest RF front-ends are moving towards multi-mode, multi-band PAs, based on the traditional technology for PAs—gallium arsenide (GaAs). The new PAs handle more frequencies, but it’s still difficult to support all 40 LTE bands; the RF front-end would end up being too big and costly. So for practical purposes, a 4G handset generally is configured with a different RF front-end to support various bands in a specific region, a sometimes complex and cumbersome process for OEMs and carriers alike.

But now there are some new options in the mix, which could help solve the band fragmentation problem for LTE and also turn the RF market upside down. One vendor, Peregrine Semiconductor, has been sampling a PA based on a variant of silicon-on-insulator (SOI) technology called silicon-on-sapphire (SOS).

And looking to accelerate the deployment of LTE, Qualcomm recently unveiled an RF front-end device, based on a mix of bulk CMOS and SOI. Instead of using an RF module, Qualcomm’s solution is housed in a package-on-package (PoP) configuration, enabling OEMs to save board space and re-configure the device more easily for a given region.

Multiple sources indicate that Qualcomm’s RF front-end incorporates the industry’s first multi-band, multi-mode PA based on SOI. Qualcomm declined to comment, saying the company isn’t ready to break out the technologies within the device. But after dissecting Qualcomm’s device, analysts said the part poses as a potential threat to GaAs-based PA suppliers, such as RF Micro Devices, Skyworks, TriQuint and others. “Qualcomm fired the first shot across the bow,” said Eric Higham, an analyst at Strategy Analytics, a research firm. “The subsystem consists of an antenna tuning IC, an envelope tracking (ET) IC for Qualcomm’s PA and a multi-mode, multi-band CMOS PA fabricated using a silicon-on-insulator substrate.”

Christopher Taylor, an analyst with Strategy Analytics, added: “This does not mean the death of GaAs, but the Qualcomm announcement undoubtedly signals faster acceptance of CMOS PAs. To stay competitive, GaAs PA suppliers will have to continue to innovate, and they may also need to offer their own CMOS PAs for the most cost-sensitive phones, as Skyworks and RFMD have already done.”

All told, there are some dramatic changes taking place in the RF front-end, where CMOS, SOI, and SOS are making inroads at the expense of GaAs. “GaAs has been displaced by SOI in the switch,” said Rodd Novak, chief marketing officer of Peregrine. “The PA is the next thing to conquer. The stranglehold that GaAs has on the power amp will start to erode.”

RF complexities for LTE
The stakes are high in the RF front-end, a $5 billion business, according to Strategy Analytics. The big market is LTE, a technology that boasts data rates of up to 100 megabits per second, which is up to 10 times faster than 3G. In total, there were 88 million connections on LTE networks in 2012, but this number is projected to jump to 322 million in 2013 and 1.6 billion by 2017, according to the firm.

LTE could grow even faster, but in many respects the technology is being held back amid a slew of challenges, namely the band-fragmentation problem. Today, there are four frequency bands in 2G cellular networks and five for 3G. “Right now, there are about 40 cellular LTE bands in total when you add 2G, 3G and 4G worldwide,” said Peter Carson, senior director of marketing for Qualcomm. “And so the challenge in terms of getting to scale in an LTE device, meaning the ability to design one device and be able to ship it anywhere, is really a function of how many bands you have in LTE.”

The problem is that many countries support their own LTE frequencies, making it difficult for handsets to provide coverage for all 40 bands. “Each country has its own frequency challenges,” said Shane Smith, vice president of mobile devices global marketing at TriQuint. “So, you are dealing with multiple bands in each country at a 3G level. This proliferates in LTE. And then with global roaming, (the bands) can’t interfere with each other. And that’s where the RF complexity is significant.”

In another example of the complexities, AT&T uses Band 17 and bought some spectrum in Band 4 for LTE. Technically, the two bands are not contiguous. But AT&T has implemented carrier aggregation techniques to make them look contiguous. “That’s the benefit and advantage of carrier aggregation, but that causes the RF architecture to change (to meet) that new requirement,” Smith said.

Generally, the 2G and 3G cell phone is relatively simple. Chipmakers ship an RF front-end, which includes a discrete PA that would support a particular band. In contrast, OEMs face some difficult choices with LTE. In theory, OEMs could build a “universal” handset that could support all LTE bands, but that could be large and expensive due in part to the RF content, screen size and other features. “You would be paying a lot of extra cost for bands that may or may not be used,” Smith said.

In a more practical scenario, OEMs can develop “regional” phones that support limited bands in a given region. But still, the question is how much RF content does a “regional” handset require? It depends on the type of handset and price point. As a rule of thumb, Smith draws the line at four bands. A handset that requires four or more bands may need multi-mode, multi-band PAs, while cheaper discrete PAs are suitable for a phone with anything less than that.

“Of all the smartphones shipped this year, the average band count is actually still less than four. Some 60% to 70% of the market would probably lean towards a more discrete solution, whether that is a discrete PA or putting two power amps in one package,” he said. “Some 30% to 40% of the market would take advantage of multi-mode, multi-band PAs. The ones shipping today would probably (support) six to seven bands. Then, on top of that, they also have discrete PAs, which can be populated or de-populated depending on the region they want to support.”

OEMs face other complex choices. To date, the PA has been dominated by GaAs. Now there are some new and emerging PAs based on CMOS, SOI and SOS, all of which promise to provide more integration and have lower power than GaAs. What’s next? “The RF antenna switch is moving from III-V materials to SOI,” said Paul Boudre, chief operating officer at Soitec. “GaAs pHEMT will not disappear, but it will remain for more specific devices.”

Soitec sees a surge in its RF business, where the company develops substrates based on bonded silicon-on-sapphire (BSOS) and high-resistivity SOI. “Our technologies’ market penetration in smartphones and other RF-based communication devices proves that our engineered substrates are competitive,” Boudre said.

GaAs is still a better solution for the PA, TriQuint’s Smith contends, but SOI still has its place. “All of the traditional RF manufacturers have SOI designers and are making many of our switches in SOI,” Smith said.  “SOI has better insertion loss and some natural linearity aspects due to the materials that GaAs pHEMT switches could not meet very easily. The SOI performance actually meets or exceeds (GaAs pHEMT). And there is a cost advantage.’’

The new contender

Leveraging the benefits of bulk CMOS and SOI, Qualcomm recently rolled out the RF360, a front-end solution that combines a PA, antenna switch, antenna matching tuner and an envelope power tracker. Supporting all seven cellular modes, the RF360 also works in conjunction with Qualcomm’s digital cell-phone chipsets.

Qualcomm integrated the PA and antenna switch into one device. “What we tried to solve here is what we call the LTE band fragmentation problem,” Qualcomm’s Carson said. “The integration of the PA and antenna switch frees up the board area so you can have enough space for the filters, duplexers and additional switches to support roaming bands, and have a single design that can be shipped to any market.”

Another key is that the device comes in a PoP package, which cuts board space by 50%. “It allows (OEMs) to have a faster development cycle,” added Steve Brown, senior director of product management at Qualcomm. “By just changing the top of the PoP package, you can actually have a different set of characteristics in bands for a given region and phone.”

The RF360 is based on both CMOS and SOI. “It’s a mix-and-match of SOI and CMOS,” Brown said. “What we’ve done is look at each of the various areas and look at the best way to get to the highest levels of integration.”

For PAs, many argue that GaAs has a huge power-added-efficiency (PAE) advantage over CMOS. Brown dismissed that notion, saying CMOS and SOI are indeed ready for LTE. “You can actually use CMOS for very complicated RF front-end solutions. For example, we have a GSM, UMTS, CDMA and LTE front-end all on one piece of silicon,” he said.

Another key to Qualcomm’s PA is a technology called envelope tracking. In this approach, the voltage is constantly adjusted to make sure the PA is operating at peak efficiency. “PA efficiency is a challenge,” Carson said. “You don’t want to waste power and generate heat. Those two things are critical to smartphone design because you want to preserve battery life. If you don’t do something like envelope tracking, you actually waste power.”

Qualcomm’s rivals are keeping a close eye on the company’s new RF solution. “Do I think it’s a competitive threat long term? Sure,” said TriQuint’s Smith. “But I also think the CMOS solutions are not superior in performance to GaAs (for the PA).”

Qualcomm already dominates the cell-phone chipset business. Many OEMs may want to differentiate their RF front-ends and not get locked into using both Qualcomm’s chipset and RF solution, Peregrine’s Novak said. In any case, Qualcomm’s solution is a step towards bringing out the long-awaited single-chip, monolithic RF front-end. But it’s unlikely that OEMs will see a single-chip RF solution anytime soon due to cost. “The RF architectures are also changing so quickly,” Novak added.

By Mark LaPedus
A survey revealed that 41% of U.S. adults who are married or in a relationship and have a computer say that time spent on the PC is a source of stress in their relationship. In fact, 59% cited work-related issues as causing relationship strain and 75% indicated that finances were a cause of stress in their relationship. The survey was conducted online by Harris Interactive on behalf of Crucial.com.

Taiwan’s United Microelectronics Corp. (UMC) continues to fall behind its competitors. UMC recently said it will move directly from 28nm to 14nm finFETs, thereby skipping the 20nm node. The new problem: UMC is having yield issues with 28nm and is behind in ramping up the technology, according to the company.

GlobalFoundries has not committed to build another new fab in upstate New York. However, the company this week filed plans for the possible construction of a new plant with 475,000 square feet of manufacturing space, according to various reports.

At the Common Platform Technology Forum, GlobalFoundries announced results from the industry’s first implementation of a dual-core ARM Cortex-A9 processor using finFET transistors.

GlobalFoundries also disclosed several of its customers at the event, including Adapteva, Cyclos and Rambus.

Also at the Common Platform Technology Forum, IBM confirmed that extreme ultraviolet (EUV) lithography will likely miss the 10nm node. Now, the industry is looking at inserting EUV at 7nm. The delays are not surprising. What’s surprising, and scary, is that many of the problems with EUV are not engineering issues. They’re due to pure physics, namely how to generate enough consistent power for the EUV source. “We’re talking about physics challenges,” said Gary Patton, vice president of IBM’s Semiconductor Research and Development Center. “This is real physics.”

Cadence announced an agreement to acquire Cosmic Circuits, a provider of analog and mixed signal intellectual property (IP) cores.  In addition, GlobalFoundries has certified Cadence’s EDA tools for custom/analog design for its 20nm LPM technology.

Soitec’s SOI wafer shipments for radio-frequency (RF) applications have increased by 400% in the last two years.  In fact, SOI is having a profound impact on RF designs and processes.

Mentor Graphics announced the next-generation of the FloEFD concurrent computational fluid dynamics (CFD) simulation product.

Fujitsu and Panasonic will merge their semiconductor units and form a new company. For some time, the system LSI businesses of Fujitsu Semiconductor and Panasonic have struggled.

Packaging and assembly are key segments of the semiconductor supply chain in China. There are more than 200 companies competing in the packaging and assembly market in China, according to SEMI.

LED bulb prices are expected to drop from $23 per 1,000 lumens in 2012 to $10 per 1,000 lumens in 2015, and then down to $5 per 1,000 lumens by 2020, according to SEMI.

The BioMEMS market is expected to grow from $1.9 billion in 2012 to $6.6 billion in 2018, according to Yole Développement. The BioMEMS market includes pressure sensors, silicon microphones, accelerometers, gyroscopes, optical MEMS and image sensors, microfluidic chips, and microdispensers.

NAND flash revenue was $19.7 billion last year, down from $21.2 billion in 2011. Revenue will pick up this year and will rise to $22.4 billion after last year’s stumble, according to IHS iSuppli.

Facing a relentless onslaught from tablets, smartphones and solid state drives (SSDs), hard disk drive (HDD) market revenue in 2013 will decline by about 12% this year, according to IHS.

Polysilicon suppliers to the solar photovoltaic (PV) industry have reduced their plant utilization rates during the past six months, with average quarterly utilization rates falling below 70%, according to Solarbuzz.

Limited commitments by touch-screen suppliers and ultra-slim panel makers are putting a squeeze on the ultra-slim PC market, according to NPD DisplaySearch.

By Mark LaPedus
The rapid shift towards smartphones and tablets is driving the need for new and low-power chips at finer geometries.

Today, the latest application processors, integrated basebands and other digital cell-phone chips are 28nm planar devices. And it won’t be long before OEMs incorporate 20nm planar and finFET devices in their systems as a means to reduce power and extend battery life.

The mobile revolution is also having a profound impact on radio frequency (RF) designs, processes and packaging. For example, there were four frequency bands in 2G cellular networks and five or so for 3G. In comparison, the next-generation, 4G wireless standard known as long-term evolution (LTE), ultimately could support 43 bands at multiple frequencies.

To support all 4G/LTE and legacy bands worldwide, the smartphone would incorporate a gigantic, power-hungry and expensive RF front-end. So, it’s difficult to envision a “world smartphone” that supports all 4G/LTE bands in every country. Instead, consumers may settle for “regional” smartphones that support some but not all 4G/LTE bands.

“The big challenge is to cover as many bands as possible in a region with an easy-to-use and low-cost system,” said Christopher Taylor, an analyst with Strategy Analytics, a market research house. “As long as the industry moves toward more bands and faster data rates, it puts more stress on the component guys.”

Responding to the demands, RF chipmakers are rolling out a new class of multi-mode, multi-band power amplifiers for 4G/LTE, based on traditional gallium arsenide (GaAs) technology. And suppliers also are ramping up new power amps based on CMOS and a silicon-on-insulator (SOI) variant called silicon-on-sapphire (SOS).

Other changes are taking place in the RF front-end. “The RF antenna switch is moving from III-V materials to SOI,” said Paul Boudre, chief operating officer at SOI wafer specialist Soitec. “GaAs pHEMT will not disappear, but it will remain for more specific devices.”

In 4G/LTE, there is also a need for a new class of diversity switches and tunable capacitors. All told, RF chipmakers are moving toward integrated RF front-end solutions in an effort to boost power efficiencies and battery life at lower costs.

Sea of RF change
In total, RF component sales are expected to grow from $22 billion in 2011 to more than $30 billion in 2016, according to Strategy Analytics. One of the big drivers for digital and analog chip makers is 4G/LTE, a technology that boasts data rates of up to 100-megabits-per-second, up to 10 times faster than 3G.

4G/LTE smartphone shipments are projected to triple from 90.9 million units in 2012 to 275 million in 2013, according to Strategy Analytics. However, the complexity of 4G/LTE smartphones is expected to come at the expense of power and battery life, thereby requiring a new class of low-power, multicore chipsets.

Broadcom, Intel, MediaTek, Qualcomm and others are shipping cell-phone chips based on bulk CMOS. Taking another approach to the problem, ST-Ericsson has rolled out an integrated cell-phone chipset based on 28nm, fully-depleted SOI (FD-SOI). The FD-SOI part is 30% faster than bulk devices, said Joel Hartmann, executive vice president of front-end manufacturing and process R&D at STMicroelectronics. “We have demonstrated a 50% power reduction,” he said.

The next breakthrough could occur by year’s end, when Intel hopes to ship its first 22nm finFET device for the mobile market. The foundries will enter the finFET market at 14nm. “Going to 14nm finFETs will help with the battery life,” said Ajit Manocha, chief executive of GlobalFoundries.

Like the digital market, there are also challenges for the RF front-end in 4G/LTE. “The current challenge is handling many of the worldwide LTE bands. The second challenge is MIMO involving multiple carriers. The third is handling the smart antennas for all bands and multiple input/output streams,” said Will Strauss, president of Forward Concepts, a research firm.

The classic RF front-end in cell phones includes three main parts: the power amplifier, antenna switch and filter. For years, the cell-phone power amp has incorporated GaAs-based heterojunction bipolar transistor (HBTs) technology. The power amp amplifies RF signals in the phone.

Typically, the RF front end is separate from the digital modem and transceiver, which are based on 28nm and 65nm/40nm CMOS or other processes, respectively. The interaction between the RF and digital blocks present some major challenges, including the ability to maintain the isolation between the frequency bands, said Thomas Richter, senior marketing director at RF specialist Skyworks Solutions.

Generally, a 2G or 3G cell phone required separate and discrete power amps to support the various bands. In 4G/LTE, cell phones must not only support LTE, but also the existing GSM, EDGE and WCDMA standards. To date, there are 17 bands in place for 4G/LTE worldwide. That list could grow to 43.

In any case, it’s impractical and too expensive to build a cell-phone that supports all 17 bands. “I don’t think anybody will fit 17 power amps in a phone,” Richter said.

The solution to the problem is the advent of a multi-band power amp, which is a single device that supports a wider frequency range. Multi-band power amps, however, suffer from a power-added-efficiency (PAE) drop, as compared to a discrete device. “Multi-band power amps make for cheaper RF solutions, but no current power amps can handle all the bands,” Forward Concepts’ Strauss said. “There are multi-band power amps in the market, but only for a few bands. Smartphones still require multiple power amp chips.”

In 4G/LTE, a smartphone may incorporate a mix-and-match of devices, possibly six discrete power amps and one multi-band power amp in the same system. Taking another approach to the problem, Skyworks recently rolled out SkyOne. This technology can combine several discrete devices, such as multi-band power amplifiers, switches, filters, and duplexing functions, in a small system-in-package (SIP). The power amp is based on GaAs, while the switch uses SOI. “With SkyOne, you can condense your PCB,” said Skyworks’ Richter.

Skyworks, RF Micro and others are also fielding CMOS-based power amps, which claim to have lower power consumptions than GaAs. “GaAs is still favored,” Forward Concepts’ Strauss said. “The only CMOS power amps have been used in low-end GSM phones in China. CMOS has not proved to be sufficient for high-performance power amps.”

The wild card is Peregrine Semiconductor, which is sampling a power amp based on its silicon-on-sapphire (SOS) technology. Peregrine’s 0.35-micron process, dubbed UltraCMOS, is a variant of SOI that makes use of an insulating dielectric sapphire substrate. SOI wafer provider Soitec provides the bonded silicon-on-sapphire (BSOS) substrates to Peregrine. “The challenge (for SOS) is getting the heat out of a power amp,” said Strategy Analytics’ Taylor.

Soitec itself is also offering a separate RF SOI technology, dubbed Wave, which are high-resistivity substrates. “These engineered substrates enable more functionality on smaller chips and lower power usage for longer battery life in portable electronics,” said Soitec’s Boudre.

Dialing up switches and tunable capacitors
IBM, TowerJazz and others provide RF SOI processes, as well. Generally, SOI, and its variants like SOS, promise the long-awaited integration of the RF front-end. But at least in the near term, however, the RF front-end will remain a collection of discrete devices.

“If you look at the front-end, you see an explosion of components,” said Rodd Novak, chief marketing officer of Peregrine. “The area for the RF front-end is shrinking. OEMs want to put more and more battery content in the system. Integration is really the driving force, as opposed to just price erosion. That’s a big change.”

Peregrine and others provide another key part of the RF front-end: the RF switch. RF switches route signals between the antenna and the handset core, through one or more signal paths. As the design of the mobile device becomes more complex, more signal paths are required.

The RF switch was once dominated by GaAs. As of late, Peregrine’s SOS technology has been winning RF switch sockets at the expense of GaAs. In response, the GaAs suppliers, RF Micro and Skyworks, are now pushing RF switches based on SOI. “We’ve displaced GaAs,” Novak said. “Now, the GaAs guys are using a highly insulated SOI substrate. But we believe sapphire is the highest insulating substrate.”

4G/LTE is also propelling a new and emerging component–tunable capacitors. These components tune the antennae to boost efficiencies. Peregrine is selling components based on SOS. Paratek, a subsidiary of Research in Motion (RIM), and STMicroelectronics, are selling components based on barium strontium titanate (BST). Another vendor, WiSpry, is offering a MEMS solution.

“There is such a huge amount of bandwidth to cover with a small antenna (in LTE),” Peregrine’s Novak said. “In fact, the industry is looking for multiple antennae now. We are trying to prevent those antennae from cross correlating. It’s almost impossible to do that without some level of tunability. ”

Tunable capacitors are now being integrated into the transceiver block. “In 2013, we are also going to see the entry of tunable networks. You will see up to three tunable components to provide a greater tuning range,” Novak said.

What’s next in RF? Some are looking at software-defined power amps. And in the distant future, tunable capacitors may end up being integrated in the power amp. “People are trying to develop multi-mode and multi-band power amps. But to do that efficiently, you need a tunable output match with tunable components,” he added.

By Mark LaPedus
Applied Materials introduced its Centura Avatar high aspect ratio etch system, capable of handling the carbon-based mask materials used to define the vertical openings in 3D NAND.

Soitec said that its Altatech subsidiary has installed a CVD system in CEA-Leti. The research organization will use it to develop phase-change memory devices and high-k metal gates for the sub-20nm node.

United Microelectronics Corp. has licensed IBM’s technology to expedite the development of the foundry’s next generation 20nm CMOS process with FinFET  transistors. This agreement between UMC and IBM is only inclusive of IBM’s 20nm CMOS and FinFET. UMC’s internally developed 20nm planar process will be aligned to IBM’s design rules and process/device targets, while UMC’s FinFET will be offered as a low-power technology enhancement option for mobile computing and communication products.

Intel announced its seventh Intel Science Technology Center (ISTC). Part of a $100 million effort to propel new R&D centers, ISTC will focus on social computing. In one of its vast array of research projects, Intel and others provided details about the development of a smart automotive headlight for seeing through rain and snow.

As expected, Micron Technology and Elpida Memory announced that the parties have signed a definitive sponsor agreement for Micron to acquire and support Elpida. The deal is worth 200 billion Yen ($2.5 billion).  In a related transaction, Micron also announced a separate agreement with Powerchip, a Taiwanese corporation, and certain of its affiliates to acquire the Powerchip group’s 24 percent share of Rexchip. That deal is worth approximately $334 million.

VLSI Research released its upbeat outlook for 2013. The semiconductor equipment market is expected to reach $56 billion in 2013, up 10.1% from 2012. VLSI Research also raised its fab tool forecast in 2012, from minus 7.2% to minus 4%.

In addition, VLSI Research said the worldwide IC market is projected to hit $283.4 billion in 2013, up 9.7% from 2012. VLSI Research also lowered its IC outlook in 2012, from 4% growth down to 2.3%. “Chipmakers are becoming more tenuous about the second half, resulting in limited visibility for equipment suppliers,” according to the research firm. “However, the skittishness has little to do with the chip market, where the overall fundamentals remain relatively healthy; it’s driven by a deteriorating macroeconomic picture. Fears are that the chip industry could be running on borrowed time given the high correlation between it and the macro economy.”

Pricing in the DRAM market has become less volatile since the February bankruptcy of Elpida Memory according to IHS iSuppli.

If ASML can get EUV to work in production, the company could hit the jackpot. An EUV tool is expected to run $125 million each. C.J. Muse, an analyst with Barclays, said: “We believe the true EUV tool demand ramp will occur in the 2014/2015 timeframe. This should coincide with DRAM transitioning to 20nm and foundry to 14nm. Based on the layer requirements, we see foundry and DRAM as the major drivers of EUV tool demand in 2015. In our base case for 2015, we are assuming (about) 18 EUV tools will go to foundries, (about) 8 to DRAM, (about) 4 to Intel, and none to NAND (though ramping in 2016), for total 2015 shipments of 30-36 EUV tools.”

Maxim announced a $200 million investment to upgrade its U.S. wafer fabs in Beaverton, Ore.; Dallas and San Antonio, Texas; and San Jose, Calif. The investment will be used for 200mm upgrades. The analog chip maker maintains a 300mm foundry deal with Powerchip.

MagnaChip now offers a new 0.18-micron bipolar-CMOS-DMOS (BCD) process. This new process features operability at 60 volts and will support additional voltage ranges (12 to 60 volts) for applications that include DC-DC converters, power-over-Ethernet, LED drivers, audio amps and power management ICs for the mobile and consumer markets.

TowerJazz said its 0.18-micron BCD process has been qualified to meet the certification requirements of the AECQ100 standard as defined by the Automotive Electronics Council (AEC).

Gaas Labs, a private investment fund, acquired Nitronex, a supplier of gallium nitride (GaN) RF chips.

Ultratech has acquired the rights to a collection of patents from IBM. These include fundamental patents in packaging such as C4 bumping, BGA, lead-free solders and 3D packaging.

By Mark LaPedus
The ongoing push towards green and energy-efficient systems is prompting the silicon foundries to jump on the bandwagon and devise their next-generation processes based on ultra-high voltage technology.

For some time, several foundries have offered 1- and 0.5-micron, ultra-high voltage processes with ratings up to 800 volts. But seeking to get a jump for the next wave of designs, the specialty foundries have taken a narrow lead in the process race over the larger players like GlobalFoundries, TSMC and UMC.

South Korea’s Dongbu HiTek and Germany’s X-Fab Silicon Foundries AG recently rolled out 0.35-micron processes with ratings at 700 volts as a means to reduce cost and power. X-Fab, for one, has moved from a 1-micron silicon-on-insulator (SOI) process to a 0.35-micron bulk technology, although the company is developing a 0.18-micron SOI scheme for 200 volt applications.

The other foundries also are working on ultra-high voltage processes at 0.35-micron and below. Ultra-high voltage processes fall into the broad category of power management and generally involve technologies from 600 volts and above. In the 600 to 800 volt segment, there is also an emerging collision course for various transistor types.

The main applications for 600 to 800 volts include AC-to-DC switching power supplies, LED lighting systems and power converters. For these systems, the market is migrating towards 0.35-micron geometries on 200mm wafers, said Thomas Hartung, vice president of marketing for X-Fab. “For analog and mixed-signal companies, we see 0.35-micron as the sweet spot,” Hartung said.

The shift towards 0.35-micron processes is expected to lower the manufacturing and product costs for systems. But the real problem has been evident for some time: How does the industry reduce or tame standby power?

In the home, for example, power supplies take AC power from a wall outlet and convert it into DC. Conventional power supplies based on older linear technology are cheap but inefficient. Appliances plugged into the wall still consume energy, or standby power, even when the product is not in use. Linear-based cell-phone chargers, for example, can consume between 0.8 to 2 Watts even when they are not connected to the phone, according to chipmaker Power Integrations Inc.

Some 5% to 15% of household electricity consumption worldwide is wasted in standby mode, according to the International Energy Agency. In the U.S. alone, standby power costs households over $5 billion in electricity a year, according to Lawrence Berkeley National Lab.

One solution to the problem is the advent of switch-mode power supplies, which are generally more efficient and expensive. For switching power supplies, the goal is to reduce costs through IC integration and finer geometries. The concept is similar for fly-back converters in LED lighting systems.

In these segments, power management chips are specified to withstand breakdown voltages at 600 volts in the event of a power surge or spike. Some vendors sell integrated power management devices at 725 volts. “You want your transistors (with ratings of at least) 600 volts to operate in the 220 AC range,” said Steve Ohr, an analyst with Gartner Inc. “You want to have tolerances at 600 volts to prevent the system from failing.”

The prevalent switching technology in these types of systems is the power MOSFET. But there are big changes within the 600 volt segment, as several transistor types are emerging in the arena. “At 600 volts, you will see a clash of the titans between a range of technologies,” Ohr said. “You will see IGBTs, bipolar, power MOSFETs, gallium-nitride (GaN) FETs, and silicon-carbide transistors.”

Foundries push ultra-high voltage
On the foundry front, there are more subtle changes taking place in the arena. In digital, most integrated device manufacturers (IDMs) have outsourced a growing percentage of their production to the foundries. In contrast, the power management IDMs tend to keep their production in-house and the foundry business is relatively smaller for high-voltage devices right now, said Robert Lineback, an analyst with IC Insights.

Still, the foundries are seeing gradual growth from an emerging crop of fabless vendors in the higher voltage segments. “For TowerJazz, 700 volts makes up approximately 20% of new power designs,” said Todd Mahlen, vice president of APAC sales and power business development for specialty foundry TowerJazz Inc. “It is a limited set of customers, compared to lower voltages. In the short term, margin numbers are better for 700 volt technology.”

The foundries do not offer a 700 volt process alone. Instead, they tend to provide a complete modular, multi-volt process, which includes low- (3.3-, 5 and 6.5 volt), medium- (20 and 30 volt) and high- and ultra-high voltage (450 and 700 volt) technologies. “6.5-, 5- and 3.3-volt are used in logic and small signal analog functions. 20 and 30 volts are ideal for gate driver voltages for high-voltage MOSFETs,” Mahlen said.

Andy Brown, vice president of foundry sales for South Korea’s MagnaChip Semiconductor, added: “People may also want 500 volts, because that is the voltage that can directly interface with an AC line. 700 volts is designed for fly-back architectures” and other systems.

Meanwhile, getting a jump on the market, Dongbu HiTek late last year claimed to offer the foundry industry’s first 700 volt process at 0.35-micron. The initial process comes without an epitaxial layer. It supports a range of voltages and maintains low on-resistance (RDSon) by using a reduced surface field technique.

Then, in May, X-Fab rolled out XU035, an 200mm, 0.35-micron process for ultra-high-voltage applications. Using bulk wafers, the process supports 20 , 40 and 700 volts with low RDSon. The total mask count ranges from 13 to 18 steps. The 0.35-micron process is ideal for integrated power-management devices, where cost and functionality are key. “The integrated solutions are becoming more and more popular,” X-Fab’s Hartung said.

Previously, X-Fab offered a 1-micron process for 650 volt applications. “That was an SOI and trench isolation architecture,” Hartung said. “But for high-volume 700 volt applications, we need bulk and 0.35-micron. It’s a cost-effective solution.”

X-Fab is developing a new 0.18-micron high-voltage process based on SOI. The applications for this process are 200 volts, which is ideal for power-over-Ethernet and ultrasound units, he said.

SOI also is getting traction in other analog and mixed-signal markets. STMicroelectronics, for example, recently rolled out a 0.16-micron, SOI-based version of its BCD process for use in medical equipment and hybrid vehicles. The process combines 1.8- and 3.3-volt logic CMOS circuits with power MOSFET transistors that can operate up to 300 volts.

“For RF, SOI is also cost-effective,” said Horacio Mendez, executive director of the SOI Industry Consortium, a group that is looking to accelerate the use of SOI in the market. In RF switch applications, SOI reduces noise and cross-talk while maintaining signal power, he said.

Two Soitec Group managers — Eric Desbonnets and Stéphane Laurent — describe how SOI wafers support RF technology development.