EDN’s 2009 Innovator and Innovations of the Year: A celebration of excellence

EDN Staff -- EDN, May 13, 2010

View as PDF

For two decades, EDN’s annual Innovation Awards program has honored the most innovative technological advances as well as the designers behind those advances. This year was no different. The judges—a team effort that involved EDN’s editorial staff and you, the readers of EDN and edn.com—faced a serious challenge to decide among closely matched and highly qualified entrants in a record 30 product and technology categories. In addition, we considered nominations for 2009’s Best Contributed Article and Innovator of the Year. Read on to learn who received top honors at this year’s ceremony, which took place April 26 in San Jose, CA.


Digital MEMS accelerometer (Hewlett-Packard) HP technology enables a new class of ultrasensitive, low-power digital MEMS (microelectromechanical- system) accelerometers. As much as 1000 times more sensitive than high-volume, commercial products, the sensors can achieve noise-density performance in the less-than-100 nano-g/√Hz range to enable dramatic improvements in data quality. HP’s large proof-mass and parallel-surface electrodes sense movement over a large dynamic range with a low noise floor. The bulk micromachining processes allow a 1-mg proof mass, resulting in the low noise floor, and the parallel-surface electrodes allow dynamic ranges of greater than 120 dB. In sensor networks, these accelerometers enable real-time data collection, management evaluation, and analysis monitoring. This information finds use in a range of safety, security, and sustainability applications, such as bridge and infrastructure health monitoring, geophysical mapping, mine exploration, and earthquake monitoring.

Analog: Converters

ADS5400 12-bit, 1G-sample/sec ADC (Texas Instruments)
The 12-bit, 1G-sample/sec ADS5400 ADC has a buffered input. It improves radar and signal-intelligence capabilities and can double the capture bandwidth of signals with 12-bit resolution in test and measurement. It offers an SNR (signal-to-noise ratio) of 59.1 dBFS (decibels relative to full-scale), SFDR (spurious- free dynamic range) of 75 dBc (decibels referenced to the carrier), and SINAD (signal, noise, and distortion) of 58 dBFS. Its input bandwidth is 200 MHz, and the user-selectable single- or dual-bus DDR LVDS (low-voltage- differential-signaling) outputs provide designers flexibility to choose between I/O speed and trace count. The device’s adjustable gain, offset, and phase ease the interleaving of two or more ADCs to create a multigigasample/sec digitizer or to balance two ADCs in an I/Q (in-phase/quadrature) receiver.

Analog: Front-end ICs

ADAS1128 current-to-digital Converter (Analog Devices)
High-slice-count CT (computed-tomography) systems can capture real-time moving images, such as a beating heart, with a high degree of accuracy and detail. The 128-channel ADAS1128 current-to-digital converter for CT scanners contains 24-bit converters that change photodiode-array signals into digital signals. The product offers 128 simultaneously sampled data-conversion channels and selectable sample rates as high as 20k samples/sec. The ADAS1128 consumes 4.5 mW/channel at full speed. It has no charge loss, multiple ranges, and charge noise as low as 0.4 fC for low-dose X-ray systems.

Analog: Signal path

FSA800 USB-accessory switch (Fairchild Semiconductor)
Fairchild’s FSA800 3-to-1 USB (Universal Serial Bus)-accessory analog switch enables USB data, audio headsets, and UART (universal- asynchronous-receiver/transmitter) data to all share a connector. The FSA800 also has charger-detection and microphone-connection functions and provides superior audiosignal integrity and overvoltage protection on the bus-voltage line for charging. Cell phones’ micro-USB connector allows both charging and data transmission, meaning that users can employ standard USB accessories, including audio headsets, chargers, and cables for USB and UART data transmission, all through one micro-USB connector.

Components

IR1168 SmartRectifier IC (International Rectifier)
The 200V IR1168 SmartRectifier secondaryside rectifier-driver IC drives two N-channel power MOSFETs that act as synchronous rectifiers in resonant-half-bridge topologies. The device also has two 200V, high-speed comparators, which differentially sense the drain-to-source voltage of the MOSFET using the device’s on-resistance as a shunt resistance to determine the polarity and level of the device currents. The device uses the Smart- Rectifier control technique, which compares the sensed voltage across the MOSFET with two negative thresholds to determine the turn-on and turn-off transitions for the device. By governing the drive level of the secondary- side MOSFETs according to these three thresholds, the IR1168 ensures accurate performance without the need of a PLL (phaselocked loop) or external timing sources. The IR1168 delivers an efficiency improvement of 1.5% for a standard 240W LCD TV switchedmode power supply, a reduction in rectifierdevice temperature of 25°C, and a reduction in board size by two-thirds.


U2723A USB source-measure unit (Agilent Technologies)
Agilent’s paperback-sized U2723A USB (Universal Serial Bus) SMU (source-measure unit) supplies ±20V voltage and 120-mA current on three channels and provides accurate current measurements to the nanoamp levels. The channels can connect in series or in parallel to achieve as much as 60V and 360 mA. With a 15-msec rise time, the device improves throughput during mass testing of semiconductor devices. It is useful in consumer electronics, semiconductors, renewable energy, aerospace and defense, and communication. The SMU tests camera modules, CMOS ICs, optical transceivers, and solar cells, and it provides design validation and characterization on semiconductor devices. It also features embedded test scripts that enable high test throughput through parallel outputs with short rise and fall times. The U2723A supports entry of 200 list modes for each embedded test script, with as many as two embedded test scripts per channel. Its plug-and-play feature simplifies data transfer through the USB interface.

Design, debug, and production test, yield analysis

Tessent YieldInsight yield-analysis tool (Mentor Graphics)
The YieldInsight analysis tool enables designers to understand and identify semiconductor yield loss from production-scan-test data, reducing the time to finding the root cause by 75 to 90%. Based on initial analysis, failure- analysis engineers use PFA (physical-failure- analysis) techniques, such as emission microscopes and nanoprobing, to identify the defects in failing devices. They then take corrective action, such as a process, design, or test-program change, to improve yield. Yield- Insight mines data with respect to designs, visualizes results consistent with yield-analysis practices, identifies systematic defects, and includes drill-down functions to effectively select die for PFA. It uses layout-aware diagnosis that provides a data set for each failing device. It needs no manufacturing-equipment data, which might be unavailable to fablesssemiconductor manufacturers. YieldInsight also uncovers previously hidden yield limiters representing the last 1 to 2% in high-volume manufacturing.

Design frameworks

1080p video-design framework (Altera)
Targeting broadcast, surveillance, videoconferencing, medical-imaging, and military-vision systems, Altera developed the industry’s only FPGA-based 1080p video-design framework, which lets designers rapidly build custom video-processing datapaths with an order-ofmagnitude productivity edge. Altera’s 1080p video-design framework allows system designers to create a custom image-formatconversion signal chain. The framework comprises a collection of IP (intellectual-property), tools, and building-block video functions and interfaces, including 1080p-quality key video IP; building-block video functions for video input, output, transport, frame-buffer interface, and runtime control; a large collection of hardware- verified reference designs; a standard open-source, low-overhead video interface; and system-level-design tools for integrating processors and memory subsystems. Altera’s 1080p video-design framework provides all the key functions that the broadcast-infrastructure systems need to implement imageformat conversion.

EDA: Back-end tools

Quartus II Version 9.1 FPGA-design tool (Altera)
Quartus II Version 9.1 offers 20% faster compilation than Version 9.0 and two- to threetimes faster compilation than the nearest competitor for high-density, 65- and 40-nm designs. The release also supports Altera’s new Cyclone IV FPGAs and the Stratix-IV E EP4SE820 FPGA. A rapid-recompilation feature permits faster small ECO (engineeringchange- order)-type design changes, reducing compilation times by 50% on average for a recompilation versus running another full compilation on the design and preserves critical timing during late design changes. Quartus II performs parallel processing in all synthesis, place-and-route, static-timing-analysis, and assembler design stages. New parallel-synthesis support significantly reduces synthesis time for designs with partitions. Version 9.1 delivers an enhanced timing-driven-synthesis feature, enabling you to improve design performance in 10% less time than the previous versions. Version 9.1’s incremental-compilation feature allows you to make changes and compile just the critical blocks until you reach timing closure, reducing compilation times by as much as 70% over a flat compilation.

EDA: Front-end analysis and synthesis tools

IC Validator in-design physical-verification tool (Synopsys)
IC Validator, Synopsys’ new in-design physical-verification tool, working with IC Compiler, allows physical-system designers to efficiently conduct sign-off-quality verification during design. It automatically detects and fixes errors in the full context of area, timing, and power, before the design closes. The end result is a DRC (design-rule-checking)-clean design that easily passes the final sign-off. IC Compiler integrates and shares a data model with IC Validator. The sign-off-quality accuracy is proven through extensive foundry qualification. The modern implementation uses hybrid polygon/edge-rule processing, is multicore-enabled, and employs a unique programmable language for concise rule creation and execution.

EDA: Front-end simulation and database tools

Virtuoso Accelerated Parallel Simulator (Cadence Design Systems)
The Cadence Virtuoso Accelerated Parallel Simulator addresses the challenges of accuracy degradation of results, excessive runtimes, and huge learning curves for setup and postprocessing. The Accelerated Parallel Simulator delivers the full accuracy of the industry reference Cadence Virtuoso Spectre circuit simulator. It improves convergence and capacity for designs with hundreds of thousands of transistors and parasitic elements, including PLLs (phase-locked loops), data converters, memory IP (intellectual property), power-management circuits, and full-chip designs. The Accelerated Parallel Simulator combines proven Cadence simulation technologies, a breakthrough advanced parallel- circuit solver, and a new multiprocessing engine that runs on computer platforms with as many as 16 cores. The product also addresses performance and capacity challenges. The result is a next-generation-performance circuit simulator with full Spice accuracy; significantly better single-thread and multithread performance; improved dc-, transient-, and RF-analysis convergence; and a larger capacity footprint.


XPort Pro Linux networking server (Lantronix)
The powerful, self-contained XPort Pro embedded networking module targets advanced applications on edge devices to enable a new class of device servers for M2M (machineto- machine) edge computing. The XPort Pro, with 16 Mbytes of flash memory and 8 Mbytes of SDRAM, handles demanding applications with the power of a high-speed, highperformance, advanced-architecture 32-bit processor. XPort Pro offers a variety of robust data-encryption and authentication options. The XPort Pro Linux software-development kit, with built in IPv6 (Internet Protocol Version 6), is an integrated embedded-hardware and -software suite that provides a validated set of Linux-based applications, an extensive software library, a board-support package, and device drivers that allow designers to create custom-tailored products. Employing the stable 2.6 Linux kernel, the software-development kit includes all of the components for building secure network-enabled products using the Linux OS. XPort Pro includes Lantronix’s patent-pending VIP (virtual-Internet Protocol) Access technology that allows for seamless integration with the ManageLinx remote-services-enablement platform, allowing access to firewall-protected equipment.

FPGAs

Spartan-6 LXT (Xilinx)
Xilinx Spartan-6 LXT FPGAs provide an optimal balance of cost, power, and performance and deliver as many as eight 3.125-Gbps GTP (gigabit transceivers/low-power) and an integrated PCIe (Peripheral Component Interconnect Express)-compatible core. Xilinx Spartan-6 LXT FPGAs are the first low-cost FPGAs fabricated on a 45-nm process. Spartan- 6 FPGAs include as many as 140,000 logic cells in a new six-input LUT (look-uptable)- logic fabric, as much as 4800 kbits of RAM, and as many as 180 DSP48A1 slices; 3.125-Gbps GTPs, typically consuming less than 150 mW; integrated PCIe-endpoint blocks, providing PCI SIG (special-interestgroup)- verified Revision 1.1, Generation 1 compliance; an integrated memory-controller block that simplifies the implementation of 800-Mbps DDR2 and DDR3 memory interfaces; and additional power saving with hibernation power-down mode.

Microcontrollers

PIC24F16KA nanoWatt XLP PIC microcontrollers (Microchip Technology)
The PIC24F16KA family of 16-bit microcontrollers features nanoWatt XLP (extreme-lowpower) technology, including sleep currents as low as 20 nA, which enables battery life as long as 20 years. The devices also integrate peripherals such as capacitive touch sensing. The PIC24F16KA family includes a watchdog- timer mode consuming power as low as 370 nA, a real-time clock and calendar mode consuming as little as 510 nA, a low-power active mode consuming as little as 8 μA, 1.8 to 3.6V operating voltage for all on-chip analog and digital peripherals, and a 32-MHz maximum speed at a battery-friendly 3V. The PIC24F16KA family achieves maximum battery life from 3V batteries. They can also run all on-chip analog and digital peripherals at a minimum voltage of 1.8V. The PIC24F16KA executes 90% of its instructions in one cycle to achieve more efficient software execution, faster task completion, and lower overall power consumption across the life of an application. The PIC24F16KA features twospeed start-up, which is a battery-friendly method of waking up from sleep and performing a task in a timely manner. It achieves this goal without creating large current drains on the battery.

MPC564XL high-reliability processor (Freescale Semiconductor)
The MPC564xL family implements a new strategy for detecting critical faults in safetyrelevant blocks and provides major test capabilities in hardware to reduce the software overhead and typical board component count. The cores, memories, crossbars, communication blocks, and peripherals all feature BIST (built-in self-test) mechanisms. The device prevents common-cause failures due to clock or voltage-supply issues. The MPC564xL family provides hardware blocks for detecting clock deviations as well as hardware monitors for main voltages. The processor’s selectable modes of operation, which both support SIL3 (Safety Integrity Level 3) and ASILD (Automotive SIL Level D) requirements, are lock-step operation, which provides a software environment for redundant processing and calculations, and independent-core, or dual-parallelmode, operation, which provides a software environment for diverse processing and calculations to increase performance or to crosscheck for reliable operation. The MPC564xL features core, crossbar, memory-protectionunit, interrupt-controller, DRAM, and software- watchdog-timer “sphere of replication.” The main benefit is the ability to detect singlepoint failures, which are typically higher-rate soft errors not only in the cores but also in key submodules of the microcontroller.

Multimedia SOCs

Armada 1000 HD media processor SOC (Marvell Semiconductor)
Marvell’s Armada 1000 SOC (system on chip), also known as the 88DE3010, targets network-connected consumer-electronics applications, such as Blu-ray players, digital media adapters, set-top boxes, and Ethernet augmented digital televisions. At its nexus are dual Sheeva PJ1 CPU cores, each operating as fast as 1.2 GHz and derived from the ARM Version 5 instruction set. The device includes a dual-channel decoder, which decodes H.264, VC-1 (Video Codec 1), MPEG-2 (Motion Pictures Experts Group 2), DIVX (Digital Video Express), and other formats. It also has a Qdeo postprocessing engine, which reduces per-pixel 3-D noise and provides 3-D deinterlacing, scaling, natural depth expansion, intelligent color remapping, and adaptive contrast enhancement. It also handles video at high-resolution frame sizes and bit rates. Graphics- and audio-processing engines and security- and system-management CPUs round out the multimedia-processing mix. Other integrated peripherals include dual SATA (serial-advanced-technology- attachment) and dual USB (Universal Serial Bus) transceivers, SDIO (secure digital input/output), SPI (serial-peripheral interface), PCIe (Peripheral Component Interconnect Express), Fast Ethernet, HDMI (high-definition multimedia interface) Version 1.3, and six video DACs.

Multiprocessing

Wolverine DSP platform (Sound Design Technologies)
The 90-nm, multicore, open-platform Wolverine targets use in hearing aids. It combines four DSPs, 11 hardware accelerators, a nonvolatile memory, and several analog functions in less than 11 mm². The chip includes filtering and FFT (fast-Fourier-transform) functions and a patented cross-point architecture. This multicore approach delivers as much as 50 MIPS of processing power at a 2-MHz system-clock rate and draws only 1 mA. By adjusting the onboard programmable clock multiplier to 12 MHz, the processing capability increases to 95 MIPS. Wolverine includes 266 kbits of program memory, 369 kbits of data memory, and 256 kbits of OTP (onetime- programmable) memory. To round out the digital functions, the chip includes several interfaces such as GPIOs (general-purpose inputs/outputs), I²S (inter-IC sound), I²C (inter-integrated circuit), SPI (serial-peripheral interface), and debugging. The analog portion of Wolverine includes two audio-rate ADCs and one audio DAC for speech processing. A third multichannel, low-speed, 9-bit ADC monitors battery voltage, temperature, and analog GPIOs. Wolverine also integrates a voltage regulator for the digital circuitry as a means of reducing digital-power consumption. The analog functions consume less than 300 μA of current. The platform comes in a compact hybrid using Sound Design’s thinStax technology.

MB88395 1394 automotive-controller IC (Fujitsu Microelectronics America)
Fujitsu’s MB88395 automotive-controller IC transmits audio and high-definition video over an IEEE 1394 FireWire in-vehicle multimedia network. The MB88395 simultaneously transmits multiple streams around the vehicle, such as images from Blu-ray DVDs, digital television, blind-spot cameras, and navigation systems. It uses an 800-Mbps physical layer and link layer. The MB88395 incorporates the DTCP (digital-transmission-content-protection) encryption protocol, which prevents unauthorized duplication, downloading, or alteration of any audio or visual media. Fujitsu’s proprietary SmartCodec compresses high-resolution video to one-quarter its original size in 2 to 3 msec. The subsequent lack of perceptible time lag means that viewers can watch the same content on front and rear monitors. Using SmartCodec compression, a 1280-pixelΧ720-line video stream from a Blu-Ray disc compresses to a rate of 249 Mbps, such that the 800-Mbps network bandwidth can transmit two channels’ worth of information.

Network, timing, and ber test

J-BERT N4903B jitter-tolerance tester (Agilent Technologies)
Agilent’s new J-BERT (jitter/bit-error-rate-tester) N4903B offers complete jitter tolerance test for characterizing emerging serial-bus interfaces, supporting both embedded and forwarded-clock devices at data rates as high as 14.2 Gbps. It allows accurate and efficient characterization of forwarded-clock devices, such as QPI (QuickPath Interconnect), HyperTransport, and memory buses. J-BERT also allows accurate receiver characterization with its unique capabilities, such as variable duty-cycle distortion on half-rate clocks, delayable jitter on clock and data signals, variable de-emphasis, and the new clock doubler accessory. J-BERT N4903B considerably reduces test-setup time and complexity, significantly increasing efficiency. Integrated and calibrated jitter sources allow generation of stress conditions for multiple serial-bus interfaces, variable output levels and rates on data and clock signals, a powerful pattern sequencer for simplified setup of training sequences, automated jitter-tolerance-test routines, and fast total-jitter and eye-measurement routines.

MSO70000 series mixed-signal oscilloscopes (Tektronix)
Tektronix’s MSO70000 series of high-performance mixed-signal oscilloscopes combines high-performance analog and digital acquisition capability to address emerging classes of mixed-signal applications from embeddedcomputing and high-speed serial interfaces to wideband RF. The MSO70000’s 16 logic channels can accommodate low-speed serial interfaces in high-performance ASICs and systems, leaving its four high-bandwidth analog channels available for critical timing analysis. The logic channels also provide the signal visibility necessary for state-specific triggering in complex systems, such as DDR3 and electronic-warfare systems requiring transient group-delay measurements. Precise, deeprecord logic-signal acquisition; a 2.5-GHz differential-logic probe with solder-in accessories; and time-aligned digital and analog trigger paths for real time fault detection set this family apart.

PCs and peripherals

X25-M mainstream solid-state drive on 34-nm flash memory (Intel)
New variants of Intel’s X25-M solid-state drives employing 34-nm-fabricated NANDflash memories deliver equivalent or better performance than their 50-nm-lithographybased predecessors at a fraction of the cost. Retail-channel prices for the 80-Gbyte X25- M variant are $225 (1000), a 60% reduction from the original $595 introduction price of one year ago. Similarly, the 160-Gbyte version sells for $440 (1000), down from $945 at introduction. Intel’s drives achieve a 25% reduction in latency and two-times-faster random- write I/O operations/sec versus 50-nm solid-state-drive precursors. They also deliver 56% better system performance versus a 5400-rpm hard-disk drive per the PC Mark Vantage benchmark utility. By delivering as many as 6600 4-kbyte write-I/O operations/ sec and as many as 35,000 read-I/O operations/ sec, the X25-M leapfrogs hard-disk drives, thereby providing for markedly faster system and application responsiveness.

LTC3108 ultra-low-voltage step-up dc/dc converter (Linear Technology)
The LTC3108 step-up dc/dc converter starts up and operates from 20-mV voltage sources, such as thermoelectric generators, thermopiles, and small solar cells. The part has selectable output voltages of 2.35, 3.3, 4.1, and 5V. It has an auxiliary low-dropout regulator output of 2.2V at 3 mA. Using a small step-up transformer, the LTC3108 provides power management for wireless-sensing and data-acquisition applications. The 2.2V lowdropout regulator powers an external microprocessor, and you can program the main power output to one of four fixed voltages to power a wireless transmitter or sensors. A second output can act as the host, and a storage capacitor provides power when the input voltage is unavailable. The LTC3108’s features quiescent current of less than 6 µA and high efficiency to ensure the fastest possible charge times of the reservoir capacitor. The LTC3108’s 3Χ4-mm DFN package or SSOP-16 and small external components provide a compact approach to energy-harvesting applications.

Power: Lighting

iW3610 dimmable LED driver (iWatt)
You can use the iWatt iW3610 as a primaryside regulated PWM (pulse-width-modulator) controller for 5 to 25W dimmable LED lamps or in a 15W dimmable-LED-driver PCB (printed-circuit board) with a footprint smaller than 30Χ75 mm in an Edison-base socket. The part achieves greater-than-80% efficiency and supports a maximum switching frequency of 200 kHz. It allows automatic reduction of LED current at high temperature. This approach also prevents overheating and increases the life of the electrolytic capacitors. The iW3610 eliminates the need for a feedback optocoupler. An intelligent algorithm ensures that the iW3610 automatically detects the presence or absence of a wall dimmer and allows 2 to 100% flicker-free dimming range.

Power: Special

Webench Visualizer (National Semiconductor)
The Webench Visualizer Web software allows engineers to see the results of multiple power-supply-design configurations. It builds on the foundation of the Webench Designer online power-supply, filter, sensor, and signalpath- development tool. Rather than analyze just one design, the Visualizer extends the analysis to every possible option. It can show design results for 25 dc/dc-power-supply architectures and more than 21,000 components from 110 manufacturers. Roughly 48 billion alternative input values are available for a single power supply to cover the possible values for input voltage, output voltage, and output current. The software highlights the design that best meets the desired dialedin weighting and shows alternative designs. Each design includes a BOM (bill-of-materials) cost for components for each socket to meet the user’s dialed-in preference.

Power supplies/systems

SolarMagic power optimizer (National Semiconductor)
In the traditional PV (photovoltaic)-system architecture, panel mismatches from realworld conditions negatively affect power output. Traditional PV systems rely on a centralized inverter; as each string of solar panels in series connects either directly or in parallel with other series strings to the inverter, the inverter can “see” this module array only as one electrical equivalent. Consequently, small amounts of mismatch can cause disproportional energy losses. To overcome the panelmismatch problem, National Semiconductor’s researchers and engineers used proprietary algorithms to enable the SolarMagic power optimizers to extract the maximum energy available by applying a localized MPP (maximum power point) and translate input voltage and current to the best output voltage and current, maximizing energy transport. The devices sense their voltages and currents and jointly adjust for them until they reach an optimum value for the string.

RFICs

Si2170 TV tuner (Silicon Labs)
The Si2170 globally compliant analog/digital-TV tuner has an analog-TV demodulator in a CMOS IC. The Si2170 eliminates the need for more than 100 discrete components and integrates a low-noise amplifier and a high-Q tracking filter that provides gain only around the desired channel frequency. The Si2170’s patented digital low-IF architecture addresses the problems of analog and digital reception and the existence of multiple regional standards.

RF/microwave test

VectorStar microwave vector-network analyzer (Anritsu)
The VectorStar microwave VNAs (vector-network analyzers) deliver best-in-class frequency coverage of 70 kHz to 70 GHz, dynamic range of 103 dB at 67 GHz, and measurement speed of 20 μsec/point. The 70-kHz, low-end device provides seven octaves of additional information below the traditional 10-MHz cutoff point of conventional microwave VNAs. For applications requiring 70-kHz to 2.5-GHz bandwidths, VectorStar uses a mixer-based receiver with bridges for directional devices. At frequencies greater than 2.5 GHz, it uses a harmonic-sampling receiver with couplers for directional devices. A VNA that operates in the frequency domain can supply timedomain information using IFFTs (inverse fast-Fourier transforms) to convert frequencydomain data to the time domain.

Sensors

AS5011 Hall-effect magnetic- encoder IC (Austriamicrosystems)
The one-chip, 2-D AS5011 Hall-effect magnetic- encoder IC from austriamicrosystems monitors a magnet’s displacement relative to its center position and provides position information through its I²C (inter-integratedcircuit) outputs. It integrates five Hall sensing elements for detecting lateral displacement as great as ±2 mm, a high-resolution 12-bit ADC, an XY-coordinate and motion-detection engine, and a smart power-management controller. The on-chip processing engine frees system designers from integrating complex software algorithms on their host processors and reduces development time. The Easy- Point joystick module provides a mechanical stack incorporating a navigation knob with a magnet and the AS5011 IC.

Silicon intellectual property

DesignWare SuperSpeed USB 3.0 IP (Synopsys)
The silicon-proven Synopsys DesignWare SuperSpeed USB (Universal Serial Bus) 3.0 IP (intellectual property) offers on-chip instantiation of the IP through certification and production test, saving cost and lowering the risk of separately acquiring and integrating the IP. The product comprises digital controllers, PHY (physical-layer) IP, verification IP, virtual-platform software, and production-test and debugging software. Innovations include support for a range of cost-effective reference clocks, asynchronous operation during at-speed loop-back testing, the ability to test receiver eye compliance after the receiver equalizer during production test on any digital tester, an on-chip receiver sampling scope to help debug signal-integrity issues, and a virtual platform for early software development.

Software/embedded tools

mbed microcontroller (ARM Holdings)
The ARM mbed microcontroller rapid-prototyping tools and developer platform enable designers to prototype with modern microcontrollers. Features of the mbed platform include a microcontroller hardware form factor suitable for prototyping, an integrated programmer that works as a USB (Universal Serial Bus) flash drive, online compiler tools that require no installation, software libraries that allow interaction with microcontroller peripherals without implementation expertise, and online developer resources that enable reuse and collaboration.

---

Innovator of the Year Numonyx Alverstone phase-change-memory design team

Current market demands are testing the limitations of flashmemory and DRAM usage in consumer products and enterprises. As increased numbers of computing devices, such as multifunction cell phones and mobile Internet devices, come to market, additional capacity is necessary to store and process increased amounts of information with less power. Additionally, as Internet-server applications offer faster online-search processing and faster response for consumers, servers must be able to handle the extreme increase in “near-memory” storage necessary for maintaining development of these applications. To address this burgeoning issue, the Numonyx researchand- development team developed Alverstone, the memory industry’s first commercially available device using PCM (phase-change-memory) technology—the most significant memory advancement in 40 years. PCM blends the best attributes commonly associated with NOR-type flash memory, NAND-type flash memory, and RAM or EEPROM. The result is a bit-alterable, nonvolatile technology with high read and write speeds and high endurance capabilities. Because physical- state change rather than electron storage powers PCM, shrinking lithographies do not present an issue. In fact, smaller lithographies are better suited for PCM because there is less material to alter between states, which dramatically decreases the power required to change a PCM bit. Consistent reliability and lower power give PCM technology an advantage over today’s memory technologies as it scales to smaller-than-45- nm manufacturing nodes, well beyond the point that you can expect NOR and NAND to reach in the future. The RAMlike performance and low-power nonvolatility of PCM offer device manufacturers a new approach to storage and higher-performance computing problems, achieving much of the promise of storage-class memory. Its scalability and lowcost structure offer a scaling path beyond NOR and NAND as these electron-storage-based memories encounter limitations. Alverstone represents the first step toward commercialization of the PCM technology and is the result of nine years of extensive research and development to bring this significantly new memory concept and storage physics to production capability.