On the fringe

Motor innovation for a greener world

2013-05-08T15:35:00Z

Globally, electric motors use nearly half of all the electric energy we produce. Let me repeat that: HALF OF ALL ELECTRICTY. And while there are over 11 billion motors produced each year – most of which are used in very low power applications like disk drives, DVD trays and small cooling fans – it is in the 100W+ motors where small efficiency gains can lead to massive cumulative savings.

In fact, one TI customer is using our latest motor control innovation in ventilation fans for an installation at a farm in The Netherlands. This customer was able to reduce total motor energy consumption by 80 percent! This is because, like many motors in use today, current was simply being wasted without producing power that was being put to use. The same is happening in a variety of variable speed, variable load motors that touch our lives every day: auxiliary pumps, compressors, and specialized traction motors in vehicles; compressors, fans, blowers, and pumps in home and building ventilation systems; refrigerators, washing machines, and dishwashers in our houses; and a variety of industrial fans, pumps, generators, compressors, and tooling that keep the infrastructure of our factories, businesses, and cities alive.

If we increased efficiency of just this set of motor-driven systems we could potentially deliver a 15 percent reduction in overall global energy consumption - that’s huge! The European Commission has been adopting directives to improve the environmental impact of these motors and drives at the design stage, and I foresee similar initiatives spreading across the globe.

The way to save massive energy is simply by enabling more engineers with the solutions that have only been available to the experts today: cost effective advanced motor control chips that can take advantage of complex current control strategies with additional solutions to the overall system design challenges faced during development. We are doing this with innovative software that we market as our InstaSPIN™ family of motor control solutions. With these solutions we put expertise on-chip. On low cost C2000 Piccolo chips to be exact, and our revolutionary FAST™ software encoder eliminates the mechanical sensor which has been such a deterrent to the deployment of highly efficient current control techniques.

We also solve system design challenges by identifying individual motor characteristics (tracking and compensating for their changes over time), automatically tuning the current control system for maximum efficiency, and enabling rapid tuning and deployment of simple to complex velocity and motion control systems. This expertise is enabling more “high use” applications, like compressors, fans, pumps, escalators, industrial drives, and traction machines, to finally be controlled in the most efficient way at an affordable system cost.

While I may be a “motor head,” I like to think I’m contributing to a greener world by helping TI pioneer some of the innovations.

Beyond CMOS

2013-04-29T18:48:00Z

Today, the smallest feature sizes on integrated circuits are approximately 20 nanometers. It’s now fairly clear that CMOS will be further scaled,at least by about another factor of ½. However, we are already well into the region of diminishing returns from this historical scaling trend. For example, microprocessor clock speeds essentially saturated almost 10 years ago. Note that this is largely due to the challenge of heat-removal, which, of course, also reflects the fact that high-speed scaled CMOS now uses relatively more power.

Texas Instruments has been a member of the Semiconductor Research Corporation (SRC), supporting university research for over 25 years. Just this year, the SRC Focus Center Research Program, now called “STARnet,” and the SRC Nanoelectronics Research Initiative (NRI), both working on “beyond CMOS,” have been successfully recompeted for the next five years. In addition, some of the STARnet centers will also address new circuit and system techniques for better exploiting CMOS.

I’m looking forward to the discovery of new devices by NRI and STARnet that might be further developed by the semiconductor industry into practical technologies for supplementing and/or replacing CMOS in many applications. Of course, CMOS is a tough act to follow. It’s even possible that there isn’t anything that could be developed over the next few decades into a practical technology that would be significantly superior to CMOS in speed, power efficiency, and cost per function. This is largely because the field-effect transistors (FETs) that serve as the “switches” in CMOS logic are so elegantly simple and effective for controlling the flow of electric current. However, as the great breadth of research in NRI and STARnet indicates, we are definitely not idea-limited!

For example, one of the research threads aims to improve on the turn-off characteristics of the switch. Below the gate threshold voltage, FET current drops by an order of magnitude for at least 60 millivolts of drop in voltage on the gate. This turn-off rate is limited by the mechanism of thermionic emission for the flow of electrons from the source to the channel of the device. Research is underway on devices in which quantum tunneling through the source-channel barrier replaces thermionic emission over the barrier. The main challenge for such “tunneling” devices is to conduct sufficient current when the device is “on” (i.e., gate voltage above the threshold). In part, this is being addressed through using new channel materials – for example, graphene as a replacement for silicon. Graphene also plays a role in research on other potential “beyond CMOS” devices, and you can read more about it in the blog by Luigi Colombo.

Another promising area of research on superior electronic switches involves materials that are called “multiferroics.” This means that they exhibit more than one of the properties of ferromagnetism, ferroelectricity, and ferroelasticity. Many of these materials are perovskite transition-metal oxides. Those combining ferromagnetism and ferroelectricity offer the promise of controlling the flow of spin/magnetization with electric charge rather than with current-generated magnetic fields, which is a much less efficient process. Thus, for example, we can envision devices in which the flow of information is via waves of electron-spin-polarization rather than simply electric current. Research is also underway on devices in which the orientation of nanomagnets represents a logic state. Such devices represent a type of nonvolatile logic that retains its state even when no power is supplied. In this case, the main challenge is in making circuits that operate at high speed. Of course, speed versus power efficiency is almost always a major trade-off, even in non-electronic technologies!

Lose your fear of robots

2013-03-19T16:40:00Z

Humans have always been the nexus of the information flow: information is gathered from the environment, analyzed, organized, communicated, and, often times, used to help guide our actions. Our need for more information has lead us to achievements such as the first transatlantic telegraph cable, which was laid in the mid-19^th century, financed with private money.

In less than 150 years later, we find ourselves in the portable, personal, wireless device and internet age. I'm not sure about you, but I am confident that I have nearly all the information I want. We are saturated by information and there is a concern about the lack of growth opportunities in the information industry. Our industry, in particular, has always been focused on information technology, so should we be worried?

No, I believe we are being joined by new consumers of information that will soon outnumber us: robots. I refer to them in the broadest sense. Robots have been around for awhile, like the ones that manage your automobile's antilock-brake system (first developed for commercial use in 1970s). Now, humans are demanding greater convenience in their everyday life, and robots are becoming more sophisticated. Some recently available robotic technology includes devices that analyze your behavior and temperature preferences, and then automatically set the temperature in the office or house, or devices that can keep the floors clean. But soon, robots will relieve us of even more sophisticated information processing tasks, like driving a car or diagnosing an illness.

The increase in capabilities of embedded processors, and the analog devices that form the connection between the real world and the digital world, is resulting in the rapid technological advances that we see today in embedded systems. These systems are more and more regarded as “robots” with access to the vast resources of the cloud.

I wonder sometimes why automation is not being adopted faster… but looking at the literature and movies featuring robots, I see a deep-seated distrust of intelligent machines. My guess is that we fear losing direct control of our own environment.

On the positive side, robots will give us a better opportunity to do more meaningful things with our lives, and in order to benefit fully, we must overcome our emotional barriers against them and embrace them instead.

Graphene’s potential to extend Moore’s Law

2013-03-01T21:03:00Z

Graphene continues to gain momentum as a potential candidate to further extend Moore’s Law for decades to come. The material shows great promise for future thermal management and energy storage devices, those that require operation at very high frequencies, and many more. TI is actively engaged in research and development initiatives to determine graphene’s full potential for advancing the future of electronics.

So what is graphene? Graphene is a monolayer of graphite that is carbon hexagonally arranged. The band structure of a single hexagonal layer of graphite was first described in 1947 by P.R. “Phil” Wallace, but it wasn’t until 2005 that graphene was isolated on a dielectric surface and its properties measured.

There have been many publications on graphene and graphene-related materials since 1947, mostly on the structure of these very thin films. By isolating graphene, researchers have successfully measured and studied its chemical, mechanical, optical and electrical properties. Graphene shows superior properties in many respects in comparison to existing materials. For example, graphene is chemically inert, impermeable to many gases, and extremely strong -- in fact it is stronger than steel! Graphene has the highest thermal conductivity of all materials, > 3000W/(m K) or a factor of about 10 higher than copper. It also has high carrier mobility, and only absorbs 2.3 percent of visible and infrared radiation. But by far, graphene’s strength and low-cost are the two biggest factors contributing to its compatibility with today’s wafer processing techniques.

Graphene’s properties make it useful for many applications, from composites to flexible electronics, sensors, thermal spreaders, touch screens, transparent conductive electrodes, photonics, high frequency electronics and others.

One of the biggest challenges that graphene faces before it can be used for most electronic applications is that, while it is ambipolar and has high carrier mobility, it does not have a bandgap in comparison to silicon’s bandgap of 1.1 eV. The graphene community is trying to develop new devices, in order to meet the beyond CMOS device requirements. Graphene has also been considered for high-frequency device applications, but it is competing with more mature III-V-based device technologies. However, III-V materials are believed to reach a limit at around 850GHz and perhaps graphene can satisfy applications beyond this.

Through TI’s collaborative research strategy and commitment to technology innovation, the company is supporting U.S. universities to develop the new switch for “beyond CMOS” through the Nanoelectronics Research Initiative (NRI). Two of the NRI centers are investigating ways to take advantage of graphene for electronic device applications. The new SRC STARnet program, of which TI is also a supporting member, is also funding activities on various applications of graphene including beyond-CMOS devices, sensors, thermal management, and interconnect.

Graphene has many salient properties that can extend silicon scaling to deliver higher performance, lower power and greater energy efficiency that customers require. It is up to us to invent exciting new ways to take advantage of them!

What’s on the horizon: the ubiquitous camera

2013-02-01T05:17:00Z

Cameras are becoming smaller, more embedded, and almost everywhere. Coupled with computing power and wireless connectivity, cameras are revolutionizing the way we acquire, understand, and share information. They are extending our eyes to places we could not otherwise see, and breaking boundaries of physical distance to connect people to bring more safety, convenience, efficiency, and fun to our lives.

On the horizon is life with the ubiquitous camera. Imagine in the morning as you drive to work, cameras help you see a 360 degree surround view of your car, eliminating blind spots and helping you drive and park more safely. Or, imagine riding in an autonomous car which relies completely on cameras, embedded processors, and wireless communication to navigate while you work, make conference calls, eat breakfast, or even take a nap!

At work the ubiquitous camera provides surveillance at your home to continuously monitor your house. When there is a suspicious event, you will automatically receive an alert, and will be able to view or playback video streams remotely. Cameras at your elderly parents’ home will constantly monitor and detect adverse events such as falls, trips, or abnormal gaits, etc. that could be indicative of a problem. You or local emergency contacts will be notified as incidents happen so prompt actions can be taken.

Wearable cameras, in the form of glasses, goggles, or a clip etc, are also enabling new and exciting capabilities, giving us more opportunities to capture and easily share the ever-elusive best moments of babies, animals and athletes in motion, and breathtaking sceneries at your blink.

On the medical front, imagine capturing and sending images to your doctor, in an effort to speed the prescreening process. In the future, tele-presence robots equipped with cameras could even allow you some face time with a remote doctor for basic diagnosis. Cameras and Internet connectivity will enable more prompt medical help and reduce much unnecessary time in waiting rooms and on commutes.

As single cameras extend into stereo cameras and camera arrays, you will enjoy your images and videos in life-like 3D effect with enhanced resolution and higher dynamic range. Coupled with computing power, cameras are becoming more like human-eyes, where they detect and capture an image, and can understand the image as well. Cameras on your game console, your glasses, watch, mobile phone, TV, doors etc will allow you to control these devices with your gestures, such as playing games, answering a phone call, changing TV stations, or locking the door. Cameras will also become part of the on-line shopping experience, allowing you to try new clothes in the virtual fitting rooms and see instant results.

Adding projector technology to cameras and computing brings a whole new experience of immersive augmented reality, where virtual objects or desired information collected by your local computer (such as an embedded processor) through analyzing the images/videos captured by the camera are then projected into the real world.

Many of these exciting camera applications already exist today, and the rest is only a matter of time. At the heart of the ubiquitous experience are TI’s analog and embedded processing technologies, such as TI’s OMAP and DM3x family of processors, along with innovative imaging, vision, and video algorithms. TI technology innovations, in analog and embedded processing -- coupled with wireless connectivity, computing, projector, display, and sensing technologies -- will fundamentally change the way we aquire, understand and share information to make our lives safer, more convenient, and more fun!

What capabilities do you hope to see from the ubiquitous camera? We’d love to hear from you to uncover the endless possibilities together!

CES 2013: Health and Fitness Buzz

2013-01-10T19:44:00Z

Besides being an electrical engineer - I’m a consumer. Yes, I’ll admit it. I’m the person that all those clever commercials successfully influence – even if I don’t want to admit how much. And if I’m completely honest I’ve come home from the store more than once with something more “featured” than I planned because of a talented sales pitch. So being a part of the CES 2013 is intoxicating. Every day I’m walking the show floor to connect with TI customers – but in the background I’m also getting a satisfying consumer high drinking in the exhibitions and always on the lookout for what gadgets I don’t know that I need yet – but will surely still need.

Health and Fitness is one of the biggest growth areas of the show this year – and it’s pretty clear why. Never before has the general population been more interested in their own personal fitness and more empowered to pay attention to bodily health. So it’s natural that there’s a surge of new companies and technologies aimed at capitalizing in this trend.

Want to count your calories burned, steps taken, and floors climbed? Done. Want to compete with your online friends on who can lose more pounds? Piece of cake. Want to analyze your blood pressure trend during that Monday morning staff meeting? Possible!

But for the eager consumer inside me, the most exciting buzz this week on the health and fitness floor is definitely trends in heart rate monitoring. Measuring heart rate during exercise is not a new concept of course, but most available solutions implement a bio-potential measurement via a chest strap. Millions of fitness enthusiasts put on their chest strap monitor every day and are relatively happy. But if you are like me and have been halfway through your Zumba class and had to stop to tug your strap back into place- Or come home from a long run only to find chaffed skin from the salty sweat collected underneath (you know what I’m talking about) – then you will be thrilled that the latest trend is to find ways to reliably measure heart rate without needing that chest strap.

And this is where the excited engineer inside me finally gets to come out and party along with my inner consumer. As part of a semiconductor IC company, it’s our technology and chips that enable this type of new technology. In our Texas Instruments Health and Fitness Zone in North Hall Meeting Room N116, one of our focus demo’s we have this year is an end- to-end solution example of a PPG based heart rate monitor that uses an LED/Photodiode based measurement to measure the blood flow in the veins of the wrist and extract heart rate data. The main measurement is done by TI’s AFE4400 analog front end that incorporates the necessary LED transmit and receive path circuitry. Behind the AFE4400 we have the MSP4305528 which takes data from the AFE and performs some motion cancellation calibration to account for body movement. And finally, a BLE module using TI’s CC2541 streams data to an iPAD real time. Companies looking to implement a strapless heart rate monitor solution can use this demo platform as a starting point for their hardware and algorithm development – and can start their design confidently knowing that a successful solution is possible. How cool is that?

So what is even better than the consumer high of finding the next big thing? It’s being a part of the technology that enables spreading it to others!

TI semiconductor IC’s can help make just about any dream possible – but it’s YOU who make the killer applications that make it necessary. So where will you take it next?

Analyzing your DNA: semiconductor sensor technologies can change the game

2012-12-17T15:48:00Z

Recently, I asked a few people about the first time they heard of DNA sequencing. The typical answer I received was, “Oh, you mean DNA analysis? I see it on CSI all the time”. A couple people even told me the first time they heard the term was during the OJ Simpson trial. No matter when you first heard the term DNA analysis, it was not the last. Why? DNA sequencing and its ramifications on our medical health is having a tremendous impact on the way medicine is progressing, and this is only going to accelerate and become and more prevalent for health care in our lifetime. It’s going to become commonplace to hear about DNA analysis for us and certainly for our children.

We will keep hearing about DNA because DNA is essential to life. DNA tells our cells what to do - From when to divide and when to die, to when to grow hair and what color. They make a human being into a unique individual. They determine our height, hair color, physical capabilities and many characteristics. Unfortunately, DNA can also determine if you will get a particular form of cancer or if you will have high cholesterol. DNA defines who and what we are, and ironically what will become of us in the future.

What is DNA? DNA stands for deoxyribonucleic acid, and it contains the coded instructions that determine what every organism is and does. So what is DNA analysis? For me, DNA analysis was a job. My first encounter with DNA sequencing was in college. I had taken a job at a university microbiology department that got me involved in cancer research. It was 1982, and the Human Genome project had not even started yet. Many cancer researchers, like the group I was working with, were looking for clues to the causes and cures for cancer in this thing called a double helix – also known as our DNA. My part in this research was to write software and to develop tools that could be used to identify specific strands of deoxyribonucleic acid or “DNA”. I knew nothing about genetics, but thankfully I knew how to write software and build some basic electronic systems I essentially helped create a system that would take the DNA that the researchers had chemically altered to fluoresce a particular color according to their DNA makeup when exposed to ultraviolet light. Further, when we applied an electric current and heated up the DNA, these strands lined up according to strand length. It allowed us to sequence our strands of DNA and determine something about their makeup. This is basically what a DNA sequencer does. There is a big fancy word for the DNA sequencing process known as GEL electrophoresis. What GEL electrophoresis really does is break apart the DNA and line up the DNA strand fragments by size. Ultimately, it allows us to measure DNA fragments to figure out what they are.

So why would we want to figure out DNA strand fragments?

The goal of the research team I was involved with was to identify DNA in plants that would inhibit cancer cell growth. In a nut shell, we wanted to discover compounds in plants that possibly could be used as an inhibitor to cancer cell growth and even allow cancer elimination altogether. The key was in identifying the patterns of DNA and seeing how they fit together. In some cases we had to modify DNA strands to see the result. Suprisingly, DNA is actually very easy to modify. So what is the process?

If you have had any exposure to digital logic, you may know that any digitized element can be represented by a “1” or “0 .” Computers often use something called hexadecimal notation to represent a digitized element. (Hexadecimal notation is just a numbering tool for representing large numbers of these bits of digital matter “1” or “0”.) As most people generally know today, a pixel of a digitized Rembrandt painting, or a clip of your favorite mp3 can be represented digitally in hexadecimal notation by the numbers 0-9 and the letters A-F.

Likewise, sequences of DNA strands are represented by one of only 4 characters: A,C,G,T, which represent something called a nucleotide. There are only 4 types of nucleotides (adenine, guanine, cytosine, and thymine). These four nucleotides are connected together via chemical bonds and form DNA chains. Four Nucleotides are all that is required to sequence an entire human. Nucleotides also pair up in a certain fashion. A is always opposite of T, G is always opposite of C. These pairings of A with T and G with C are called base pairs. There are 3 billion base pairs in the human genome

As you can see, knowing how DNA strands pair up and that there are patterns help us recognize and identify DNA. These 2 characteristics mean that if a portion of a DNA strand is identified and part of it is missing, we can pair both the available and missing sequence against many other pieces of DNA to see if we have a MATCH. We can also use a piece of DNA strand to make a duplicate strand, called cloning. The DNA of every human is both; almost identical, and very different.

How can that be? Well, we all have 23 chromosomes. Our DNA is organized into about 3 billion bases on those chromosomes. We can identify whether or not DNA comes from a human, a horse, or a dog, or a plant. However, if we look closer, we can map out specific sequences of DNA and give a person specific characteristics. Although much of your DNA might match other humans, you also have specific DNA that makes you “you”.

Below is a diagram of DNA sequences that have been exposed to UV Light. Accompanied by increasing strands of nucleotides…

Due to similarities in DNA strands, the more DNA we have, the more we can narrow down to a smaller and smaller group of people with a specific DNA code – leading eventually to an individual.

However, since there are billions of DNA base pair codes for humans, we would need a pretty sophisticated machine to sequence all of the DNA for a complex organism like a human being. The Human Genome Project decided to take on this task. The Human Genome project broke the process up into pieces. Teams of people worked on identifying the genetic code of humans in sections. On April 14, 2003, it was announced that the Human Genome was complete. Did this mean that the work was done? Obviously not. It meant we now had a reference that we could explore. The human genome provided adecoder ring that could help us discover what makes us “US.”

So how does the human genome relate to the world of semiconductors and technology? Well, now that the human genome DNA sequence has been identified, it can be searched for specific characteristics. For instance, if you want to know your potential risk for a disease like high cholesterol, we can run your DNA and find whether you have a “marker” - a particular sequence of DNA that has been identified with many others who have that condition. You could then take precautionary measures to ensure you minimize that risk. In the future we might find that we can develop gene therapy – modifying your DNA to eliminate that marker and turn off that DNA characteristic so that the symptom, or prohibit a disease so it never takes effect.

The ramifications of DNA sequencing are immense both inside and outside medicine. Today, researches are modifying the genetic makeup of eColi to make it produce hydrocarbons that could be used for fuel. Even crops are modified to make them resistant to certain diseases.

All of this sequencing of DNA has been done a sophisticated instrument called a DNA sequencer. DNA sequencers allow many strands of DNA to be identified and tested in a very short time, essentiallygiving any individual the chance to have their own DNA mapped. Unfortunately today this equipment is still expensive and complicated. $10,000.00 isn’t an uncommon price to pay for a DNA sequencer.However, this is changing rapidly. Right now there is talk about a person being able to get their entire genome sequenced and analyzed for around $1000 and eventually below. A reduced price would allow the medical community to use DNA sequencing in ways that are just too expensive today.

The driving factor for this capability is improvement in the expense and capabilities of DNA sequencing instruments. Developing cheaper instruments is one step towards this goal. Just think about the transition that computers went through from the 50’s until now. Remember what happened to allow desktop computers today to emerge in the 1980s from the room size, or even table size, computers of the 1980s. The genetics industry is going through a similar process. Geneticists want cheaper and more capable sequencing tools. Technology being employed in the world of DNA analysis is poised to drive the same kind of results that would emulate the changes in the PC industry.

Inexpensive electronic components is just one way that the price of DNA sequencers can be reduced, but it’s not the only way that semiconductor technology from companies like Texas Instruments could apply their technical expertise.

Now, more than ever, semiconductor technology is playing a larger role than just making cheap electronics. Some genetics companies are transitioning from the optical detection methods described above to electromechanical detection methods. A key objective now on sequencers is to run many streams of DNA sequences through a series of parallel small protein nanopores and detect them electrically. Specialized CMOS semiconductor technology is now being employed to detect nucleotides directly. For example, miniaturized pH sensors are now being developed using ion sensitive field effect transistors (ISFETS). These pH sensors work by using a silicon nitride gate insulator that can quickly transduce hydrogen ions into a voltage, thus, allowing scalability for parallel streams of nucleotide evaluation.

What these really are are solid state biosensors developed using CMOS semiconductor technology. Semiconductor technology is emerging now as a key element for helping this industry to increase its capability and reduce its cost. Very similar to what semiconductors did for the electronic industry over the last few decades. Semiconductor technology is now playing a very direct role in the analysis of DNA sequencers in the form of sensors. it appears that the many of the companies developing this biosensing technology are fables semiconductor solutions providers – making CMOS semiconductor-based biosensors a potential opportunity for companies like TI with extensive fab potential. In any event, the worlds of bio-technology and semiconductor technology are becoming increasingly linked, and have to potential to provide mutual benefits to each other in the future.

A picture of DNA sequencers today. But there are desktop sized DNA sequencers that are just becoming available.

IEDM recap: Achieving ultra-low power in next gen applications

2012-12-11T17:14:00Z

Earlier this week, we had the pleasure of giving an invited talk at the IEEE International Electron Devices Meeting (IEDM) in San Francisco on ultra-low power design and future device interactions. This was on work that I had done with Ajith Amerasekera, about 600 people attended the presentation and as always we found the event to be very engaging, informative and fun, with many good discussions with industry colleagues after the session!

This event has a history dating back nearly 60 years, with a focus on technological breakthroughs that promise to shape the future of electronics. And I can’t think of an area more important to the future of our industry than innovation in ultra-low power and energy efficiency. As semiconductor technology continues to drive more intelligence in systems and continued growth of the Internet of Things, the ability to develop system-level solutions for energy management, delivery and consumption is crucial. At the foundation of these system-level solutions are ultra- low power and mixed signal circuits and ICs, where innovation must occur to advance applications and fuel growth.

In our presentation, we looked at some of the areas where process technology can enable ultra-low power design, including:

Performance ICs – Today’s ICs are pushing more performance and mixed signal integration into smaller areas with low power and low cost. As the CMOS process paradigm shifts from performance and geometry scaling, to one that is aimed at meeting power system needs, the focus on co-design of process and circuits for advanced integration will grow. Typical SoC chips do not have a uniform performance and power demand and usually have a variation in performance/power demand by block that can be fixed or vary through time/function, based on application demands. Process technologies need to provide a range of knobs to allow design/architecture optimization through that space using power supply, transistor width, transistor gate lengths, threshold voltages, back-biasing, and other similar techniques.

CMOS scaling and low power –Various techniques to adapt dynamic power and leakage power by changing transistor design parameters will enable us to continue delivering performance and power benefits in future circuits using CMOS scaling to our advantage. We also don't necessarily need continuous optimization points across the performance/power range, but we do need big bang for the buck with low complexity/cost. This is a complex trade-off between design optimization granularity, chip requirements and silicon process technology. It is especially important if it can enable cost competitive and easier manufacturing. For example, three threshold voltages not ten, three gate lengths not sixteen. Even with the supply voltage, we may opt to use three and not have too much fine grain control with little extra return on technology cost. We may not need to use transistor widths that are granular down to the minimum design grid; a selection of widths and transistor pitches can work well if they are selected correctly. These have to be co-optimized with circuit design and layout, and the device designers in silicon manufacturing development.

Interconnect advancements – As mixed signal ICs more to more advanced process technology nodes, and are integrated into complex SoCs with kilometers of wiring, transport delays can have a negative impact on both performance and power. Solving challenges of wire scaling is a must-do in building future system architectures for ultra-low power. Metal system metrics must cover the system. For instance, both short distance low fan-out for logic clusters, and long distance low fan-out for block-block communication must be comprehended. Clocks, resets, enables, etc. are also issues that need to be addressed.

Ultra-low power design – Various approaches to address ULP and high performance on the same process node will continue in next generation IC design, including methodologies that optimize for power where performance is not critical, and maximize performance in small areas of the chip where performance is required. New techniques such as multi-chip modules and 3D stacking will also further balancing power and performance. Watchdog or wake-up circuitry and the need for high performance analog RF is also important. There is a push for 'battery free' or a 10-year battery, scavenge sensors, and leading-edge technologies may be off limits due to power and cost issues. We need to leverage both new and scaled technologies, in addition to learning from existing and older technologies to provide the eventual solutions needed by these products.

The bottom line is that future applications depend on important architectural and system-level changes that will be driven by advancements at the process and chip level. At TI, we take a systems-level approach to low power and energy efficiency across the entire signal chain. It includes all aspects of system design, from product development and design and the processes that serve as their foundation, to packaging and manufacturing and software development. Check out our website to learn more.

My Definition of Innovation

2012-12-07T17:34:00Z

I was listening to the radio on the way to work the other day and heard an interesting segment that started my thought process of defining the difference the difference between creativity and innovation. I don’t remember exactly how much of the segment I am borrowing and how much I have created but here is the result of my being innovative on the topic. So here it is in the form of four statements.

Knowledge: Knowing the right answer to the question.

Intelligence: Knowing the right question to ask.

Creativity: Asking the question for which there is no answer.

Innovation: Answering that question.

While I am at it, let me add a fifth definition to the list: “Business is making money with the answer.”

As engineers we sometimes get accused of not being innovative. But I disagree with this accusation. Rather we should be accused of not being creative enough. We are taught to be troubleshooters from our first day in engineering school. That gives us the ability to figure out the answer to any question (problem) we are asked. Here is where the difficulty occurs: Who is asking the questions that don’t yet have answers?

So, that is a good question. Who is responsible for asking the creative question? Several years ago we invited Dr. Anantha Chandrakasan to visit Texas Instruments to talk to us about his research. Upon his arrival I warned him that one of our engineers was certain that what he and his PhD students were doing was impossible. His simple reply was “you know, unless it is impossible it is not research.” If the first attempt to answer the question doesn’t appear to be impossible then it may not be very creative.

Once the questions appears to be answerable, the development process begins, and it turns into a team sport. From the point of the question being asked and a solution is ready for prime time, there are hundreds of innovations that must happen.

So, are you asking the impossible question? Or are you trying to answer it?

The Tip of the Spear: TI and Innovation

2012-11-26T16:16:00Z

Texas Instruments is not only a well-known name in the semiconductor industry it is also a well known name in the engineering world. I’m sure quite a few engineers (especially the newer ones) have TI to thank for help passing their calculus exams by way of a TI calculator. Beyond the fundamentals of math and physics found in the classroom, it’s the engineers TI hires that are the dreamers of the dream. It’s those that see not only what the current market demands, but those that create and drive new markets before they even exist.

Enter products like the TMP006 temperature sensor (Take a look!). TI’s TMP006 is a contactless temperature sensor that uses an infrared MEMS thermopile to look at an area directly in front of the sensor and read back temperature data accurate within 1 degree Celsius all within a ridiculously small area of 1.2mm². This product has raised the bar to a whole new level. It pulls together vital features that serve existing markets, and unveil new markets not yet materialized in the front of analysts and investors. So what does this mean? Imagine that instead of having to run wires inside a computing system and getting peripheral temperature readings, now you could mount a device directly on top of say an applications processor and run accurate and efficient thermal management .

Too easy? Imagine integrating this sensor into your phone, making another facet in the ever evolving “swiss army” phone? Feeling a fever? No Sweat. Check your vitals. Something fishy with your sushi? Check and make sure it’s not room temperature. Done. Contactless temperature may not be new, but integrating it in a new package is blazing new trails into undiscovered territory. It’s that thrill of the chase into the unknown that thankfully excites and invigorates those that are on the tip of the spear--your friends at TI.

Video Platform Video Management Video Solutions Video Player

The future of biometrics technology

2012-11-07T13:17:00Z

We’ve all heard the saying, “The future is in the palm of your hand.” With the proliferation of biometric technology, it has never been truer. The future really is in the palm of your hand, but did you know it’s also in the pattern of your iris, the minutiae of your fingerprint, and the structure of your face?

What does this mean to us and the future? Imagine walking straight through airport security without waiting in line or being able to quickly and securely get money from an ATM without having to remember a personal identification number (PIN.) Picture accessing your workplace without using an identification or access card. These innovations are quickly becoming realities as our world shifts from using access cards and PINs to the new digital age of secure identification and access control based on biometrics.

The global biometrics technology market is growing at a compound annual growth rate of about 21.6 percent, with the iris, vein and face markets leading the growth, according to MarketsandMarkets’ Global Biometrics Technology Market (2010-2015) report.

This biometric technology is making access control and identification systems more accurate and secure. PINs, passwords and the like are easily hacked, and cards and badges are easily stolen. These access control mechanisms are based on what we know or possess, which can be easily compromised to steal our identities or access our homes or bank accounts. However, biometric identification accurately analyzes a unique biometric feature that is nearly impossible to replicate. For example, an iris recognition system can analyze an individual’s unique iris pattern and compare it against a database of biometric information to match (or disprove) someone’s identity. Access control and identification systems that integrate this technology are experiencing reduces levels of fraud and are easier to access.

Digital signal processing is at the core (pun intended!) of biometric-based access control and identification applications. First and foremost, DSPs enable real-time analysis of computationally intensive functions. This enables biometric systems to compare against thousands of stored user identities in a second or less. DSPs are also easily programmable, enabling OEMs to customize and differentiate their system and to update their system as increasingly cutting edge algorithms become available. Last but not least, TI’s DSPs offer a variety of security features to safeguard proprietary algorithms and sensitive user data. Need more information on our biometric technology offerings? Check out our new white paper .

TI has a variety of DSPs ideal for biometric applications, from the TMS320C5515 ultra-low-power DSP to the power-efficient TMS320C6748 DSP. Members of the TI Design Network also offer some great software and tools for these applications, including a new IriShield™ iris recognition module from IriTech. The module is powered by TI’s C6748 DSP, which accelerates computationally intensive functions in the module to enable real-time iris recognition. The DSP enables IriTech’s embedded algorithms to complete a matching query against 1,000 stored user identities in just 750 milliseconds.

Our team at TI is excited about where biometric-based access control and identification applications will take us in the future. We’re already seeing them on smartphones and laptops, at the mall and at the gym, but where will we see them next? Leave us a note and let us know the new places you’re seeing biometrics-based access control and identification applications in your daily life.

It has been Amazing

2012-10-22T14:21:00Z

I am approaching my 39^th anniversary at Texas Instruments this coming January. Not long after that I will be retiring from my career at TI. I always amazed at the response I get from people when I tell them I’ve been at the same company for all these many years. I can describe their amazement in one of two ways:

-What, couldn’t he find a better job?

-And, they still haven’t fired him, go figure.

For those of you who have had one of these reactions to my length of tenure at TI, let me give a bit of counseling.

When I hired in to the Calculator division on January 14, 1974, I remember having the same thought about finding a better job as many of you have had. I had assumed that I would stay at TI for 3 or 4 years before moving on to a better job. At the time we had colored badges to indicate the number of years an employee had been at TI. I was amazed at the number of colleagues who had blue (five years), Yellow (10 years) and Silver (15 years) badges, not to mention the real dead wood with gold badges. Were these people so bad they couldn’t find another job anywhere?

Looking back I was half right on my only staying 3 to 5 years. It was that long before I already had my third (or was it fourth) job at TI. I don’t know that I have ever kept a job for more than 5 years while at TI: each new job was better (or at least different) than the last one. I guess you might say I’ve been too busy having fun at TI to take the time to look outside of TI for that next job.

“Fine”, you might say, “but why haven’t you been fired yet?” Well, I have been, many times. My greatest career advances were a result of my being fired. I guess, a more polite way of putting it, is that I’ve been “allowed to find another opportunity within the company.” This may surprise you, but let me put it in better perspective.

I have a talent and passion that shouldn’t fit well into the mainstream of a manufacturing company. I like to start things but don’t really like being part of an ongoing business. I tell people that I get fired when two words come up in conversation: profit and schedule. But this is where TI has been an amazing company to work for. Once fired, I was always encouraged to move on to another opportunity to apply my talents in a passionate way.

If you are now wondering what that talent is, I’ll tell you. I know how to find new business opportunities that have never existed before, nurture them, and support them with the best IC’s in the industry. While doing this, I build a relationship of trust and integrity with them so that when my replacement is introduced to them, the start-up company feels comfortable with TI, ready to move forward. And, my reward? Another opportunity to get fired.

Just for the record, let me insert a short list of some of the jobs from which I’ve been fired:

Speech Learning Aids
DSP Business
Wireless business
DSL and Cable modems
MP3 and DSC
Digital Radio
Automotive vision
3D imaging

Yes, there is an interesting story behind each one of these. And, if I did my job correctly, most people who read this never knew I had anything to do with them. But I am okay with that. It’s how I roll.

Does this mean I am going away this time for good? Well, at least I’m not getting fired, right? But, my going away for good all depends on my future success as I continue to follow my passion – looking for that next big thing that technology can make us totally dependent on. I do have a backlog of the next set of big things. And, just as in the past, those next big things won’t happen without the technology that TI is known for. No, you’ll not be getting rid of me that easily.

Throwback Thursday: The Win-Win Effect of Innovation

2012-09-27T14:09:00Z

I have spent most of my career in the world of digital signal processing at Texas Instruments. What have continually amazed me over those 38 years are the remarkable innovations our customers create based on our DSP devices. Before I share a few stories about our innovative customers, let me tell another story to set them up.

Every time we have introduced a new DSP with significantly more capability, I have been asked to predict what new innovations our customers will create. Each time my prediction has been the same (and a correct prediction, I might add). I suggest that I can certainly offer an idea of some of the evolutionary things that will be done, but the ones that excite me most are those which will once again blindside us. They are innovations we can’t predict, and the new DSP wasn’t designed to do them.

I know that the term “blindsided” doesn’t sound very flattering for us at TI. But I hold it as a badge of honor. It means that every time we provide a technical breakthrough in our devices (such as higher performance or lower power), our customers’ innovation exceeds our imagination of what is possible. That is a win for us. We also create a win for our customers in that our innovation opens a door for them to walk through. It gives them an opportunity to innovate on top of our innovation. In effect, we create a “win-win” situation every time we innovate. Now I would like to tell you a couple of stories about where I’ve witnessed this scenario over the 30-year history of our DSPs.

When we began the DSP business unit, it was our practice as the DSP application and marketing team to have lunch together almost every day. At lunch we would discuss what our customers were doing. One day one of the engineers said he had been reading an article in a magazine about how a racecar manufacturer was winning Grand Prix races as a result of a new concept called “active suspension.” The article said that the system was driven by a thing called a DSP. We had a great discussion on it and decided we needed to know whose DSP they were using, as we had not been in contact with the company or anyone else on the topic and assumed it wasn’t ours. To our surprise, it was a TI DSP.

When the whole idea of a digital cell phone was first being discussed in Europe, our business team in the U.S. didn’t know about it or why anyone would want a digital cell phone. But, fortunately, our European team knew and encouraged us to make it a priority. The point they made that got my attention was that every aspect of the GSM (European Digital Cell Phone) standard was defined by how many of our TMS320C25 DSPs it took to power it. Obviously, we were designed in. Now all we had to do was to make sure we stayed designed in.

TI’s TMS320C25 DSP

Finally, the way we became the leading supplier in a new concept called Voice over Internet Protocol (VoIP) was totally a surprise to us. In fact, the first time our business team heard about it was at a brainstorming session with our field application team to see what was new. At that meeting we learned that in this new market segment we had a massive market share without us even knowing. I remember hearing that the word on the street was that we had designed a DSP device specifically for that market. We were brilliant and didn’t even know it.

So, these are a few of the many stories of the users of our innovations, innovating beyond our wildest expectations. And yes, 30 years later, it is still happening! We see it most often in two areas: 1) the perpetual device I have spoken about previously (ultra-low power) and 2) multicore processors (ultra-high performance). Both give designers using our DSPs new doors to open. I’d love to hear from you about your story of innovation on top of our innovation at TI, so leave a comment below.

------------------------------

Gene Frantz is Texas Instruments' Principal Fellow and is regarded as "the father of DSP" by many in the industry. He has been intimately involved with the growth of the technology - from theory, to products, and now its phase as a true catalyst for new markets and products. He is often quoted in the media on the direction and future opportunities enabled by signal processing.

Can Wireless Technologies Improve Health Care?

2012-09-13T15:47:00Z

As the number of baby boomers entering the health care system begins to grow exponentially, finding enough physicians to provide the same level of care we enjoy today could present some challenges. While many states are increasing the number of medical students they admit, research is showing that fewer medical students are interested in becoming the general practitioners the system may need. There is also an anticipated lack of physical facilities for the 60 million Americans anticipated to enter the health care system over the next few years.

There are several trends that are anticipated to address these problems. Preventative screening technologies are the first line of defense and are anticipated to enable an automated form of triage that starts with the patient. Retail clinic disease monitoring technologies are anticipated to further reduce visits to the hospital by helping patients that already have a chronic disease manage it more effectively. Medical tourism coupled with post operative monitoring technologies are anticipated to offload many patients who can afford to travel abroad.

Preventative screening devices are designed to trigger communication with a physician when some biometric abnormality occurs. These wireless devices, such as heart, oximeter, and blood pressure monitors are typically worn periodically to determine whether biometrics are within normal ranges. When a normalized threshold is crossed, the device alerts the consumer of the abnormality. These devices may help reduce the number of “worried well” patients visiting a doctor or undergoing unwarranted clinical tests.

Retail clinics are increasingly found co-located with pharmacies. Because a general practitioner can sign off on prescriptions through a nurse and there are plenty of pharmacies with the required facilities, this trend is anticipated to address many chronic disease management needs. Wireless devices associated with chronic disease management are anticipated to make regular visits to retail clinics more time and cost efficient for both the patient and the clinician.

Many people are also anticipated to get an increasing number of special procedures in a foreign country, and many physicians advertising medical tourism also practice medicine in the US. Here again, wireless technologies are anticipated to play an important role for postoperative monitoring, where a procedure conducted in one country may have a local follow up in the US triggered by a wireless device.

In addition to these home-monitoring technologies, wireless technologies in non-life critical hospital settings are anticipated to improve care taker efficiency and make patients more comfortable. If you’ve ever had to stay in the hospital after a procedure and turned over to wake up to the sound of loud beeping alarms, you can understand how wireless technologies might make your stay better while also reducing the number of times the nurse has to visit your room.

Click here to here for insight on how TI is innovating in these areas.

Future Friday: A traffic-free world

2012-09-07T18:43:00Z

On the way to work this morning, I was stuck in bumper-to-bumper traffic. Minutes were ticking by, and I was getting increasingly impatient, thinking to myself, “I have meetings to go to! I don’t have time to sit on the highway!” This is not a common experience for me as I normally zip by the traffic jam in the HOV lane. You see, I’m one of those lucky ones that get to carpool with his family almost everyday. As sweet as it sounds, I always worry that one of those frustrated drivers on the regular lanes could jump into the HOV lane at an inopportune time. With the large speed differential, chances for a serious accident are large.

I’m in R&D at Texas Instruments and spend much of my time thinking about what the future of the world could look like with the semiconductors that TI innovates. This morning’s traffic jam got me thinking about a world without traffic jams, or even better, accident free. What a crazy idea!

Or is it?

Intelligent systems are here today, and they could make this traffic-free Utopia possible. Intelligent systems take a lot of data, process it and produce an action, sometimes without human intervention. We see this intelligence in robots that can go from one room to another; in industrial equipment that detects defective products and removes them from a conveyor belt; and yes, in automobiles. Advanced driver assistance systems (ADAS) in cars can “see” a child behind your car as you’re backing up and automatically hit the brakes; they can detect drowsy drivers and those drifting into another lane and immediately jolt the driver back to consciousness. All of these intelligent systems are becoming more human-like in their ability to analyze a situation and make a decision as a result. What’s the secret ingredient to this intelligence? It’s what we at TI call “embedded analytics.”

Embedded analytics has several requirements. The system must adapt to different applications and increase in sophistication over time, so programmability is a must. The system must also be able to process in real time. For instance, a back-up camera on a car is much more useful when it can make a split-second decision to brake if an obstacle or pedestrian is behind it. And lastly, the system must be able to process very complex, math-intensive algorithms in order to perform analytics and make decisions. TI’s digital signal processors (DSPs) fit the bill for these needs.

Intelligent systems are here today, but they haven’t pervaded the market yet. They are about to take a huge leap in both capabilities and accessibility to the general public.

Take ADAS as an example. Right now, there are a limited number of vehicles with automatic cruise control, automatic parallel park and pedestrian detection systems. However, this market is expected to explode over the next few years and will soon be considered a “must have” feature, similar to keyless entry or automatic windows.

Our intelligent cars will communicate with drivers, to other cars, and even to infrastructure and the cloud. They will have complete information about the driver, and its environment in order to maintain a safe distance between vehicles, wake us up if we are getting sleepy, and turn or decelerate for us if we are distracted at the wheel. I’ll be able to enjoy a worry-free commute down the HOV lane. Never mind, we won’t need HOV lanes. We’re going to save more than the extra minutes it takes to get to work when a car accident occurs. We’re going to save something much more valuable—lives!

ADAS is one of the many intelligent systems that will improve our lives in the years to come. Leave a note and let us know how intelligent systems are making your life smarter, safer and more fun.