The Kynix Blog

TroublesAs the deployment of Industrial IoT systems continues to proliferate,the streams of data transferred to the cloud skyrockets, drastically increasing the cost for cloud computing.  SolutionIn order to meet this trouble, many systems designers are adopting edge computing,in which data processing is done close to the source like sensors in a bid to reduce data transfer,storage and processing costs,plus address a few other concerns over Cloud Computing,in particular security. What is Big DataBig Data is a broad label for the growing amount of data generated by IoT devices and smart systems. For instance, some aircraft engines have more than 5,000 elements that are monitored at relatively high sample rates. Most of the data is transferred to a ground station for the real-time monitoring of the engine and for future R&D work. But this is only part of a growing trend. Most ‘smart’ systems produce vast amounts of data which needs to be processed immediately or be stored for subsequent processing. Huge datacentres are required if you want to store Big Data.Big Data is a broad label for the growing amount of data generated by IoT devices and smart systems. For instance, some aircraft engines have more than 5,000 elements that are monitored at relatively high sample rates. Most of the data is transferred to a ground station for the real-time monitoring of the engine and for future R&D work. But this is only part of a growing trend. Most ‘smart’ systems produce vast amounts of data which needs to be processed immediately or be stored for subsequent processing. Cloud Computing's advantages and disadvantagesCloud Computing has a lot of advantages including cost efficiency (i.e. no need to invest in and maintain your own hardware), scalability, resource availability (for all your users irrespective of their geographic locations), lower latency (as you can specify servers that are closest to the relevant users/customers) and peace of mind in terms of back-ups. There are,however,some disadvantages also. The biggest of which is that no provider can guarantee 100% availability. Data security and privacy are also causes for concern, both on the cloud and for data in transit. Latency can be an issue for Big Data, and doing computationally intensive tasks on the cloud will increase the cost. Of these concerns the last two, in particular, can be negated through edge processing; i.e. performing much of the computationally intensive work near the source data. Benefits here include real-time or near-real-time data processing and reduced network traffic, as you need only transfer the product of the edge processing, thus resulting in lower Cloud Computing costs. Security and privacy can be improved by keeping the sensitive data (a.k.a. Hot Data) within the edge processing environment and only sending less sensitive (Cold) data to the cloud. FPGAs have the edgeThere are technologies that can be used for edge processing applications. These include the use of traditional CPUs (scoring high in terms of flexibility), application-specific processors (e.g. GPUs) and ASICs/SoCs (scoring high on performance). However, it is FPGAs that are slotting into most edge processing applications. Why is this so? Well, let’s consider the requirements. Edge processing needs to be high-performance and in this respect an FPGA can perform several different tasks in parallel. For example, consider executing many non-dependent computations (such as A=B+C, D=E+F and G=H+I). On a CPU, these would have to be performed sequentially, with each sum requiring a few clock cycles. In an FPGA, an array of adders could do the computations in parallel, possibly requiring only a single clock cycle. Power efficiency is essential too, as the end product may well be battery-powered. With an FPGA the function (design) need be the only circuit present, whereas the architecture of a CPU or GPU may not be fully utilized. Also, with an FPGA comes the benefit of reprogrammability. Higher security is afforded too because the edge processing functions are hard wired into the FPGA. It is also possible to encrypt the transaction bus and to even go as far as designing your own processor. ConnectingA prime example of where edge processing is extremely useful, and in which FPGAs can play a significant role, is within an embedded system in which data derived from images needs to be transferred. For example, in the automotive sector Advanced Driver Assistance Systems (ADAS) are under development to make driving safer, easier and more comfortable, and ADAS is regarded as a significant step towards fully autonomous cars. The data processed by an ADAS can be used to notify the driver of problems or to automatically trigger responses such as deceleration, braking and/or the execution of a manoeuvre. The data can also be useful outside the vehicle. Let's discuss the embedded vision system first though by considering an ADAS demo unit that was built for this year's Embedded Vision show in Santa Clara, California. The demo comprised a TySOM-2-7Z100 prototyping board (see figure 1) which includes a Xilinx Zynq XC7Z100 device and a TySOM-FMC-ADAS daughter board to interface with four 960 x 540 pixel cameras. The processing was shared between a dual-core ARM Cortex-A9 processor and FPGA logic (both of which reside within the Zynq device) and began with frame grabbing images from the cameras and applying an edge detection algorithm (‘edge’ here in the sense of physical edges, such as objects, lane markings etc.). This is a computational-intensive task because of the pixel-level computations being applied (i.e. more than 2 million pixels). To perform this task on the ARM CPU a frame rate of only 3 per second could have been realised, whereas in the FPGA 27.5 fps was achieved.This picture is a TySOM-2-7Z100 prototyping board. Mixed technology (like CPU and FPGA) boards are proving very popular for edge processing applications and for connecting with the cloud.  The ARM CPU was mainly used for superimposing detected edges over the initial camera images, colour-space conversions, the formation of a composite image (see main image) and outputting it to an HD buffer. The FPGA and CPU could also work together to recognise and distinguish between obstacles and pedestrians close to the car and to provide lane departure warnings. What goes upSending the processed data to the cloud for further processing and/or storage is then a relatively simple task. Firstly, an AWS account would be created along with an AWS IoT environment. Next, we would configure a Thing (seeing as it is the IoT) and download the public and private keys needed for secure communications with the cloud.The embedded C MQTT standard would be the ideal Software Development Kit (SDK), because it is secure and requires minimal bandwidth. An application would then be prepared to run on the ARM CPU to publish the data onto the cloud. Imagine a scenario,howevber,under which we have data from thousands of vehicles going to the cloud.Analysis of the data could be performed on the cloud and made available for traffic systems or highway maintenance organisations, for example. There may also be instances where data from the cloud feeds into an edge-processing application, in which case applications are also available from AWS. All in all,there are both advantages and disadvantages associated with cloud computing. And many of the disadvantages will be overcome though edge-processing that FPGAs are a particularly suitable activity.  Article provided by Farhad Fallah,an Application Engineer with EDA company AldecArticle edited by Kynix 

SummaryThe robot comes perfectly if they can walk more naturally for humans.However,it's not an easy job for robots and their designers. Walking on two legs is actually a complicated task,requiring several muscles to perform delicate balancing acts.That's why in spite of years of major technological advancements in the field,humanoid robots are still far from being able to get around easily and reliably. Engineers at EPFL's Biorobotics Laboratory are testing new walking algoritms on a plateform called COMAN, short for COmpliant HuMANoid. This 95-cm-tall humanoid is designed specifically for studying walking – which is why it has no head. COMAN was developed under the EU AMARSi project and is being used by several research teams. The EPFL team is looking specifically at the "brains" of the machine. "We developed algorithms that can improve the robot's balance while it's walking," says Hamed Razavi, a researcher scientist at the Biorobotics Lab.  Body One:Climbing Stairs and Opening Doors The algorithms are geared towards three types of realworld applications. The first is carrying out rescue missions in disastrous scenarios. "In environments designed by humans - like a nuclear power plant where there are stairs to climb and doors to open – humanoid robots can get around more easily than robots with wheels," says Razavi. The second is helping with tasks like carrying heavy boxes or moving objects (see box). And the third is creating exoskeletons for the disabled. "Making the robots more stable is just the tip of the iceberg," says Razavi. The next step is refining the algorithms so that the humanoids have a wider range of movement and can overcome obstacles and walk on irregular or sloped surfaces. Two:In Harmony with Symmetries One of COMAN's distinguishing features is its joints,which are integrated with elastic elements that give it greater flexibility when performing different tasks.The EPFL team came up with a novel control algorithm for the robot, based on the existing symmetries in the structure and dynamics of the robot' as well as the mathematical equations representing the robot dynamics. "You could say we're working in harmony with these symmetries rather than against them. As a result, we obtain a more natural and robust walking gait," says Razavi. The control algorithm uses sophisticated computer programs to carefully analyze the date received from the robot – including its position, velocity, joint angles, etc. – and sends appropriate commands to the motors, telling them what to do in order to maintain the robot's balance. "For example, if someone pushes COMAN, for example, our algorithms will calculate exactly where its foot should land in order to counteract the perturbation," says Razavi. Three: Humanoids Helping Humans As part of this project,Jessica Lanini and Hamed Razavi studied how two people carrying an object together are able to walk,turn and speed up in a coordinated manner - without communicating with each other.Their findings,recently published in PLOS ONE,indicate that the two people automatically synchronize their steps, like a quadruped. Now the researchers plan to apply their results to humanoid robots. Lanini explained:"Whether for manufacturing or natural disasters, we need robots that can interact with humans and help us carry heavy objects,but such robots don't exist. That's because, in order to operate safely and effectively, the robots would need to be able to make decisions and respond to unexpected circumstances." But such robots don't exist. The reason is that in order to operate safely and effectively, the robots would need to be able to make decisions and respond to unexpected circumstances." Article provided by  Ecole Polytechnique Federale de Lausanne. 

A while back,MEMS and Sensors Executive Congress that many designers,researchers and industry reoresentatives argued for putting MEMS devices such as accelerometers and microphones, and a wide variety of other sensors in just about everything was held in San Jose,Calif.. We heard about an electric snowboard with traction control, voice-controlled garbage cans, and accelerometers placed on the nose to listen for speech in noisy environments.But sometimes the simplest example is the most memorable. In this case, that was a MEMS accelerometer—like the one in your step-counter—that thwarts car thieves.  "Passive keyless entry (PKE) systems can be made more secure with an inexpensive accelerometer." Lars Reger, chief technology officer for NXP's automotive division said,"PKE systems unlock a car—and allow it to start with a button push—by recognizing when the “key” is close to the car, either right next to it or inside of it. This is convenient for drivers, who don’t have to remove the key from a pocket or purse. But PKEs are ridiculously easy to hack—at least when a car is sitting in a driveway and the owner is at home." This hack which demoed by Swiss researchers in 2011 and still being used by car thieves around the world today,works because most people toss car keys in a basket or on a counter fairly close to their front door—close enough that a thief with a radio outside can pick up signals from the key. An accomplice with another radio stands near the front door of the car to pick up signals from the key and transmit those signals to the car. The system, concluding that the key is nearby, unlocks the car. Earlier this year, researchers pulled off the hack with US $22 worth of gear in a demo at a security conference reported in Wired. The team suggested that changing the timing of the calls and responses from the car and key could address the problem. At the sametime, PKE key holders were advised to keep their car keys in their refrigerators, whose metal exteriors would block the key's signals. NXP put forward another solution--Enter the accelerometer (and, of course, the company is bringing it to market soon, which is why representatives are willing to talk about it). Business development manager Marc Osajda told me that NXP had initially been working on a mechanical switch to turn the radio in the PKE key on and off. Then, after the company merged with Freescale Semiconductor in 2015, engineers at Freescale made the case for using a MEMS device instead, arguing that its lower power consumption made it a good fit for a gadget with an expected battery life of a year or more.    The 50-cent component works on the assumption that if your PKE key has been sitting in one place for a while, you aren’t going anywhere, so it can turn off its radio and the microcontroller that was listening to the car’s radio; it will turn back on as soon as you pick it up. Osajda said that instead of reducing battery life, putting an accelerometer in the PKE key ends up extending battery life, because the accelerometer uses far less power than the parts it is allowing the key to turn off. It's not a easy work. Osajda said "mostly because car keys take a lot more abuse than wrist wearables". He also He pointed out that people frequently drop their keys (sometimes even out of second story windows onto concrete) or toss them into washing machines (not a surprise for keys designed to stay in your pocket). According to Osajda,NXP's MEMS switch is going into production and they will be incorporated into PKE keys from a variety of manufacturers during 2018. 

A device that turns light into sound has allowed researchers to capture lightning in a bottle, in a sense, slowing down the light beams enough so that they can be easily stored and manipulated. Researchers at the University of Sydney in Australia, have figured out how to turn a light wave into a sound wave, creating an acoustic memory that they say will help data centers save energy by eliminating some electrical connections between processors. They reported their work in a recent issue of Nature Communications. “Our vision is to replace the electronic interconnects between different processors and computing machines with photonic ‘wires,’’’ said Birgit Stiller, a postdoctoral researcher who led the project. “So light transmission will be used instead of electronic connections.” The team built a chip that consists of a spiral-shaped waveguide made from a soft glass called chalcogenide, sandwiched between two stiffer pieces of silica glass. As a light beam travels through the chip, it is met by another pulse of light that has a slightly different frequency. The difference between the frequencies of the two light beams is a “beat,” a wave with a frequency 100,000 times lower, thus turning the light wave into a sound wave.  The sound wave lives for a brief time—several nanoseconds—in the spiral chalcogenide waveguide. To read it out, the device reverses the process, adding the beat frequency to a light pulse to recreate the original light wave. In standard optical fibers, light waves are prevented from leaking out of the fiber by a difference in refractive index between the core of the fiber and the cladding wrapped around it. In a similar way, the two types of glass keep the sound wave in place; the speed of sound is much slower in the chalcogenide than in the silica. Slowing down the waves provides time to synchronize different signals coming from different processors. That eliminates the need to convert the optical signal to an electronic signal. Electronics can produce excess heat and require more energy, which are important issues in the big data centers owned by Google, Amazon, or Microsoft, Stiller says. Further work with the design and materials might allow the sound waves to be stored longer, although the memory already lasts long enough for the use they envision. She and her team hope to refine the work further, with an eye to building a prototype of a manufacturable chip within the next few years. 

Article provided by University of Illinois at Urbana-ChampaignArticle edit by kynix A new methord to make Green LEDs more brighter and more efficient have been developed by researchers who at the University of lllinois at Urbana Champaign.  Researchers have created gallium nitride(GaN) cubic crystals grown on a silicon substrate that are capable of producing powerful green light for advanced solid-state lighting. "This work is very revolutionary as it paves the way for novel green wavelength emitters that can target advanced solid-state lighting on a scalable CMOS-silicon platform by exploiting the new material, cubic gallium nitride," said Can Bayram, an assistant professor of electrical and computer engineering at Illinois who first began investigating this material while at IBM T.J. Watson Research Center several years ago. "The union of solid-state lighting with sensing (e.g. detection) and networking (e.g. communication) to enable smart (i.e. responsive and adaptive) visible lighting, is further poised to revolutionize how we utilize light. And CMOS-compatible LEDs can facilitate fast, efficient, low-power, and multi-functional technology solutions with less of a footprint and at an ever more affordable device price point for these applications."  GaN was formed ethier hexagonal or cubic typically. HExagonal GaN is thermodynamically stable and is by far the more conventional form of the semiconductor. However, hexagonal GaN is prone to a phenomenon known as polarization, where an internal electric field separates the negatively charged electrons and positively charged holes, preventing them from combining, which, in turn, diminishes the light output efficiency. Until now, the only way researchers were able to make cubic GaN was to use molecular beam epitaxy, a very expensive and slow crystal growth method when compared to the widely used metal-organic chemical vapor deposition (MOCVD) method that Bayram used. Bayram and his graduate student Richard Liu made the cubic GaN by using lithography and isotropic etching to create a U-shaped groove on Si (100). This non-conducting layer essentially served as a boundary that shapes the hexagonal material into cubic form. "Our cubic GaN does not have an internal electric field that separates the charge carriers—the holes and electrons," explained Liu. "So, they can overlap and when that happens, the electrons and holes combine faster to produce light."  At the end, Bayram and Liu still believe their cubic GaN method may lead to LEDs free from the "droop" phenomenon that has plagued the LED industry for years. For LED lighting color like green, blue, or ultra-violet LEDs , their light-emission efficiency declines as more current is injected, which is characterized as "droop." "Our work suggests polarization plays an important role in the droop, pushing the electrons and holes away from each other, particularly under low-injection current densities," said Liu, who was the first author of the paper, ""Maximizing Cubic Phase Gallium Nitride Surface Coverage on Nano-patterned Silicon (100)", appearing Applied Physics Letters. Having better performing green LEDs will open up new avenues for LEDs in general solid-state lighting. For example, these LEDs will provide energy savings by generating white light through a color mixing approach. Other advanced applications include ultra-parallel LED connectivity through phosphor-free green LEDs, underwater communications, and biotechnology such as optogenetics and migraine treatment. Having better performing green LEDs will open up new avenues for LEDs in general solid-state lighting. For example, these LEDs will provide energy savings by generating white light through a color mixing approach. Other advanced applications include ultra-parallel LED connectivity through phosphor-free green LEDs, underwater communications, and biotechnology such as optogenetics and migraine treatment. 

An entirely new model of the way electrons are briefly trapped and released in tiny electronic devices suggests that a long-accepted, industry-wide view is just plain wrong about the way these captured electrons affect the behavior of hardware components such as flash memory cells. The model which devised by scientists at the National Institute of Standards and Technology (abbreviation NIST),a measurement standards laboratory, and a non-regulatory agency of the United States Department of Commerce and its mission is to promote innovation and industrial competitiveness, was test to explain how electron capture and emission creates the insidious nosise that increasingly threatens performance as electronic devices continue to shrink in size.   KIN Cheung, NIST researcher Kin Cheung also the lead author of a new report in IEEE Transactions on Electron Devices said "Such a burst noise,popcorn noise or random telegraph noise(abbreviation RTN) have become a major problem for extremely small devicess". Charge trapping is one of the known causes of flash memory failure. The new model, which NIST physicist John Kramar called "a major paradigm shift in charge-trapping modeling," could lead to a different approach to manage this problem, and potentially, a new way of making the memory cells smaller. John Kramar explained:" Charge trapping is one of the known causes of flash memory failure,the new model whicl I called it a major paradigm shift in charge-trapping modeling,could lead to a different approach to manage this problem,and potentially,a new way of making the memory cells smaller.  What is RTN noise? RTN noise consists of abrupt random drops in voltage or current caused by itinerant electrons that are briefly captured from, and then rejoin, the main flow along a current channel in, for example, a common type of transistor called a MOSFET. "The effect was mostly negligible back in the good old days when devices were larger and there were lots of electrons flowing around," Cheung said. But in today's advanced devices, with feature dimensions in the range of 10 nanometers (nm, billionths of a meter) or less, the active area is so small that it can be swamped by a single trapped charge. "As you get down to the very smallest sizes, RTN can be nearly 100 percent as strong as the signal you're trying to measure," Cheung said. "In those conditions, reliability disappears." In the case of RTN, the basics are known: The noise is caused by the action of electrons near the interface between two materials such as an insulator layer and the bulk of the semiconductor in a transistor. Specifically, an electron is pulled out of the current flow and trapped in a defect in the insulator; after a short time, it is emitted back into the main current in the semiconductor. What actually happens on the atomic scale at each stage of the process, however, is incompletely understood. The orthodox approach to account for those effects is to treat all the trapped electrons as a single 2-D sheet of charge that extends uniformly across the center of the insulator. Each emitted electron is thought to return to the semiconductor in a reverse of the same process by which it was captured, causing very little change in the presumably stable state along the insulator/semiconductor boundary. The model is suitable for very small devices,however,it didn't make sense to the NIST scientists. Among other difficulties, it ignored the fact that, once they are immobilized, electrons cause considerable distortions in local electrical field conditions along the boundary, affecting current flow. "We're saying the traditional way doesn't really work," Cheung said. "You have to rethink this thing. The old model doesn't make reasonable assumptions about how charge carriers behave." The researchers proposed a new model, based on local effects, in which the mechanisms of capture and emission are dramatically different from the standard picture. For one thing, they determined that quantum mechanics, the modern theory that describes the behavior of these systems, makes it hugely improbable, if not impossible, for electrons to get out of the insulator the same way they got in. "It's like a highway where there is an exit ramp, but there's no on ramp," says NIST co-author Jason Campbell. "You can go in, but you can't come back that way. You've got to come back a different way. That is, there is a set of rules for capture that don't apply to emission." "When you realize that the capture and emission processes are decoupled," Cheung added, "you quickly have a very different view of the problem."  The standard RTN picture supposes a weak interaction of trapped charge with its local surroundings―in this case, the highly separated electric charge in the silicon dioxide that often makes up the insulator layer in a transistor. NIST scientists found that a weak interaction is inconsistent with known physics and not in agreement with reports from two independent laboratories. Indeed, the interaction energy of a captured electron can be more than 10 times greater than previously believed. Recognition of this stronger interaction energy enables the new local field picture to explain RTN naturally. The success of the new model, and the resulting drastic change in the understanding of both capture and emission, suggested that many long-held ideas would have to be thoroughly reconsidered. "It's a very scary and very unsetting conclusion,I mean,this is tear-up-the textbook stuff." Campbell said. As an end, NIST researchers hope the new model will help chip engineers and designers understand in much greater detail how devices degrade and hat will be required to get to the next stage of miniaturization while maintaining reliability and reducing noise.