Circuit Cellar recently sat down with several professors, engineers and one master’s student to discuss the fields of computer applications and embedded hardware and software.
Q&A: Embedded Today
Ayse Kivilcim Coskun
Assistant Professor | Boston University | Boston, MA, USA
Tell us about your research.
My main research area is energy efficient computing. I work in the general area of computer engineering, with specific focus on embedded systems, computer architecture, design automation, and software.
What do you say when a non-engineer asks: “What is a microcontroller? Is it the chip in my PC?”
A microcontroller is a chip that is designed to perform a small, restricted set of tasks. Some examples would be a chip that controls the settings of a microwave oven, a chip that controls an automated gate in a parking lot, or a chip that controls the ABS in a car. The chip in your PC is called a microprocessor. In contrast to performing a small set of specific tasks, a microprocessor is able to run a large set of “general-purpose” tasks. Your PC, for example, can run web applications, text editing tools, videos, and many other tasks.
Do you think “8 bits dead”?
Lots of application domains are making use of 32-/64-bit processors today, and 32-/62-bit processors dominate the market. Still, I don’t think 8-bit is dead. Small battery-operated embedded devices such as wireless sensors typically need long battery life, which motivates using smaller, lower-power controller chips where possible. We see a growing number of sensory devices in many diverse fields (environmental monitoring, medical devices, smart buildings, etc.), which again motivates using small, low-cost controllers for simple tasks.
What are you currently working on?
One of my main projects is using 3-D stacking technology, where multiple chips are stacked on top of each other, for improving energy efficiency of computing. There are lots of interesting opportunities and challenges. For example, we can put DRAM and processor cores in the same chip using 3-D stacking technology to cut down the “memory bottleneck.” Or we can stack more functionality into a 3-D stacked chip while keeping the individual chip area small, which is better for achieving higher yield in manufacturing. However, 3-D stacking brings many challenges, including higher temperatures on chips, lack of mature design and validation tools, and technology challenges. My research goal is to explore these opportunities and challenges for designing methods that can utilize 3-D stacking for getting higher performance out of our chips at a lower energy cost.
What’s the best engineering related advice you’ve received or given?
I think Feynman’s quote, “For a successful technology, reality must take precedence over public relations, for nature cannot be fooled” (from Space Shuttle Challenger Inquiry), is one of the best pieces of engineering advice. We often see some products or ideas out there that may be well-marketed, but they really do not have the robustness, quality, or functionality that should be there. In the long run, though, I do believe better technology, better design, and ideas win as we cannot change the rules of physics or, in other words, “fool nature.”
Do you have a go-to MCU?
Not really. My selection of products varies depending on the projects, and I try to diversify my choices a bit while experimenting with new ideas.
What was the best course, lecture, or webinar you’ve attended?
I have attended a lot of great talks, hard to select one. A non-technical talk that pops up in my mind: I really liked Berkeley Professor David Patterson’s talk on “How to Have a Bad Career in Research/Academia.” I attended it while I was switching from being a PhD student to a professor. It really gave a lot of good insight and it was fun! I recommend it to all graduate students. The slides are on the web (www.cs.berkeley.edu/~pattrsn).
Your favorite engineering pioneer?
Two names pop up in my mind. One is Feynman, as I just quoted him in a question above. He is categorized as a physicist rather than an engineer but the boundaries between the two are rather thin when it comes to experimental work. I read Feynman’s Six Easy Pieces in my freshmen year in college, which inspired me to learn how things work and build things myself to solve problems. The second name I want to mention is Grace Hopper. She was a female scientist/engineer at a time when there were even fewer women in engineering. I admire her courage, many scientific contributions, and I like the fact that she popularized the term “debugging” (which was motivated by removing an actual moth from a computer system).
What are your goals for 2013?
My research group is working on reducing energy consumption of computers – or, in other words, improving “energy efficiency” – through jointly optimizing the hardware and the software. We will continue innovating in this area, and I hope we will demonstrate improvements in computing clusters as well as in small embedded devices.
1. Print magazine or digital? Print
2. Laptop or desktop? Laptop
3. iOS or Android? Android
4. Analog or digital? Digital
5. Dealing with hardware issues or software bugs? While researching, I focus more on HW issues (such as crafting a new architecture). In daily-life computing, I more commonly deal with SW bugs.
6. IE, Chrome, Firefox, or other? Firefox.
7. Starting a project or finishing it? Starting!
8. QWERTY keypad or touch screen? Both!
9. Flash drive or cloud? Cloud
10. Webinars or onsite lectures? Lectures
Is resurrecting a Neanderthal possible?
The Huffington Post recently featured a video segment discussing the controversial remarks from George Church, a geneticist from Harvard University, about how cloning a Neanderthal could happen.
In response to Church’s comment, Assistant Professor Douglas Densmore (ECE) joined several scientific panelists who offered their viewpoints to the news website. Densmore was asked to engage in the discussion for two reasons: 1) His role in synthetic biology involves creating bio-design automation tools, software, and algorithms, and 2) Church and Densmore are members of the Synthetic Biology Engineering Research Center (SynBERC).
Neanderthals are a subspecies of human beings and have been extinct for the past 33,000 years. As a result, cloning a Neanderthal raises scientific and ethical implications.
Densmore understands that Church’s comments were taken out of context and that no one is seriously considering cloning a Neanderthal. However he does believe that as time passes, ethical standards adapt to societal values. In the video interview, he states, “If we learn more about biology and what is possible, ethical discussions will need to take place and positions will change over time. Ultimately it will be ethically imperative to use that knowledge for the benefit of society.”
Many scientists are not opposed to genetically modifying plants or animals. However, the speakers on the panel were in agreement over not crossing “human ethical standards.”
Very few synthetic biologists actually discuss cloning Neanderthals in their daily lives. Most of their laboratory conversations revolve around addressing “disease in third world countries, bio-energy concerns, and making and using cheaper materials,” Densmore said. The main focus of his team’s research, he said, is improving problems in the world.
There was a consensus among the panel that increased dialogue between scientists and the general public would lessen confusion about the goals of geneticists’ research. Finding cures to epidemic issues may not always be as headline-grabbing as resurrecting a Neanderthal, but the effects on humankind would be greater.
-Chelsea Hermond (SMG ’15)
When it comes to synthetic biology, an area of research that combines engineering and science, there are still a number of questions that need to be answered when it comes to finding out not only how life works but also how it can be used to benefit society.
Still, a growing number of scientists and engineers see the field’s potential, particularly in applications like cloning, virus detection, nanotechnology, and other areas.
Assistant Professor Douglas Densmore (ECE) is one of the researchers paving new roads in synthetic biology – and his work earned him a lot of attention in 2012.
“Media attention for synthetic biology is definitely growing,” said Densmore, who recently noticed even Fidelity Investments promoting the field in a banner at Boston’s Logan Airport. “It’s here to stay.”
In July 2012, The Scientist’s Amber Dance wrote about emerging research in the field and featured Densmore for his program, Clotho, used to manipulate DNA sequences. In the article, Densmore compared Clotho to an iPhone – his group provides apps to customize the sequencing experience and users can also write their own. Clotho now offers more than 30 apps for molecular biologists.
“More and more researchers are getting involved,” said Densmore. “People are definitely starting to see its potential.”
Earlier in the year, Densmore and Associate Professor Soha Hassoun, Tufts University, served as guest editors of the magazine, IEEE Design & Test of Computers in a special May/June 2012 issue about synthetic biology.
“While there are many questions that the larger science and engineering community must still answer regarding synthetic biology, [electronic design automation] professionals can highly contribute to this nascent field,” Densmore and Hassoun wrote.
In that same magazine, Densmore and Hassoun outlined some of their approaches in “Design Automation for Synthetic Biological Systems.”
“Computational methods and tools to (re-)engineer and synthesize biological systems, referred [to as] bio-design automation, are poised to play a critical role in the development of novel biological systems similarly to how electronic design automation (EDA) transformed designing VLSI circuits since the advent of silicon transistors in the 1950s,” said Densmore and Hassoun.
If that wasn’t enough coverage, Densmore was also featured in the August 17 issue of ACS Synthetic Biology when the publication featured two of his team’s research papers and asked him to write an editorial. He also spoke in the magazine’s podcast.
“I believe that we have roughly two choices,” Densmore wrote in the editorial. “First, we wait until biology is fully understood to create design tools, or second, we start today with what we know and create flexible, adaptive software that paves the way to a more fully understood future while providing tangible, experimentally verifiable results today.”
As interest in synthetic biology grows, so, it seems, does the interest in Densmore’s work.
-Rachel Harrington (email@example.com)
Read the ACS Synthetic Biology research papers:
“An End-to-End Workflow for Engineering of Biological Networks from High-Level Specifications”
“Automated Selection of Synthetic Biology Parts for Genetic Regulatory Networks”
Article Highlights Innovative Biomedical Optics Technique
Professor Jerome Mertz (BME), who also holds appointments in the Physics and Electrical & Computer Engineering departments, has developed a light scattering technique that could enable clinicians to look under the surface of tissue in the colon, lung, brain or other organs without relying on fluorescence labeling or any other kind of probe. Because the technique works even with thick or opaque tissue, scientists can use it to observe tissue in vivo rather than extracting samples.
Recent Boston University graduate, Robert Winnett (ECE ’12), may be new to the job market, but that didn’t keep him from being featured in an Intel Free Press article titled, “Companies Battle for Next Tech Superstars.”
Winnett’s photo, taken by Senior Lecturer Babak Kia Montazam (ECE) during senior design project testing, appears in the story which offers advice about how technology companies can draw the best talent and students can make themselves attractive employment candidates.
In the picture, Winnett is working on software used to build his team’s robot, capable of retrieving ping pong balls and returning them to a specific location.
The caption points out that graduates from BU’s Department of Electrical & Computer Engineering often find internships at companies like IBM, Raytheon, Intel, Genzyme, Microsoft, and HP.
According to Intel Free Press and Georgetown University research, technology companies may see a shortage of as many as 3 million US high-skills workers by 2018, so the demand for engineering talent will only increase over the next few years.
-Rachel Harrington (firstname.lastname@example.org)
As the need for faster computers grows, current technologies are unable to provide solutions that are sustainable.
Boston University Assistant Professor Ayse Coskun (ECE) is researching energy efficient 3-D stacked systems to help solve this problem.
Coskun’s Q&A in the July issue of Circuit Cellar is an example of the computer engineering community’s rising interest in her research on energy efficiency and thermal challenges.
Coskun joined BU’s Department of Electrical & Computer Engineering in Fall 2009 and has taught computer engineering classes including Introduction to Embedded Systems and Introduction to Software Engineering since Spring 2010.
This March, Coskun received a $450,000 grant from the National Science Foundation (NSF) for early career development and her research project, “3D Stacked Systems for Energy-Efficient Computing: Innovative Strategies in Modeling and Runtime Management.” In May 2011, she was awarded the A. Richard Newton Graduate Scholarship at the Design Automation Conference (DAC) for her project, “3-D Systems for Low-Power High-Performance Computing.”
Circuit Cellar believes her research “could change the way electrical engineers and chip manufacturers think about energy efficiency for years to come.”
Her research focuses on software optimization for green computing; thermal modeling and management of 3D stack architectures; and design and runtime management of many core systems. Coskun is an affiliated member of the Center for Information and Systems Engineering (CISE) and Center for Computational Science at BU. She is also a junior fellow at the new Hariri Computing Institute.
-Sneha Dasgupta (COM ’13)
Professor W. Clem Karl (ECE, BME) has been elected editor-in-chief of IEEE Transactions on Image Processing, the leading journal of image and video processing research.
“With image processing playing such a big role in the world, it’s an exciting time to be editor-in-chief,” said Karl, who previously served as associate editor of the publication.
Karl is an active member of the Institute of Electrical and Electronics Engineers (IEEE), a professional association for electrical engineers that boasts more than 400,000 members in 160 countries.
“I believe in giving back to the professional community and am looking forward to interacting with my colleagues and learning new things,” said Karl of his new position, a three-year appointment that begins in January 2013.
In addition to his latest role, Karl is also a member of the Board of Governors of the IEEE Signal Processing Society, member of the IEEE Bio Signal and Image Processing Technical Committee, and member of the Steering Committee for IEEE Transactions on Medical Imaging.
Karl’s research at Boston University has focused on statistical signal processing; inverse problems; biomedical signal and image processing; multidimensional signal and image processing; and synthetic aperture radar. In 2000, he was awarded the department’s ECE Award for Excellence in Teaching.
-Rachel Harrington (email@example.com)
When it comes to disease diagnosis, early detection can save lives. Unfortunately, with the current technology available, patients can sometimes wait days to find out the results.
Cue Boston University’s Assistant Professor Hatice Altug (ECE) and her research team who are developing a portable biosensor that has the potential to diagnose everything from food poisoning to swine flu in as little as 30 minutes.
Altug has already received national attention for her work. In September, Popular Science included her in the magazine’s “Brilliant 10.” Then in October, she earned the Presidential Early Career Award for Science and Engineers from President Barack Obama. Now, the Institute of Electrical and Electronics Engineers (IEEE) have recognized her contributions as well by profiling her in their latest issue of The Institute.
Altug has been working on this research since 2007 when she joined BU’s Department of Electrical & Computer Engineering and helped found the Laboratory of Integrated Nanophotonics and Biosensing Systems.
The U.S. National Institutes of Health recognized the potential of her work by giving Altug and her collaborators a $4.8 million grant to develop her sensor for commercial use.
Altug was profiled by the IEEE not only for being an accomplished researcher but also for her work as a mentor to current and prospective engineering students. She told The Institute that she “wants to give students a taste of what attracted her originally to engineering: a penchant for building useful tools and gadgets.”
-Rachel Harrington (firstname.lastname@example.org)
11/14/11 – Microarray Analysis Research by Altug’s Team Makes the Cover of Lab on a Chip
10/26/11 – Photonics Goes Flexible
9/28/11 – Altug Receives Early Presidential Career Award
9/15/11 – Popular Science Names Altug to Brilliant 10
7/25/11 – Seeing Biomolecules with the Naked Eye
6/28/11 – Altug Selected to Participate in National Academy of Engineering Symposium
12/3/10 – Hatice Altug Wins Photonics Society Young Investigator Award
Have you ever been stuck scanning aisles at the grocery store and struggling to find that last item on your list?
BU alumni, Dan Ryan (ECE ’10) and Aaron Ganick (ECE ’10), are working to solve this problem and more through their company, ByteLight, a Cambridge, Mass., startup recently featured in MIT’s Technology Review.
Read the article.
ByteLight aims to use LED bulbs as indoor positioning tools that can be used in applications to help people navigate and offer deals targeted to a person’s location.
Ryan and Ganick hope that department stores will replace some of their traditional bulbs with Bytelight’s design, expected to be available within a year. Their fast-flickering lights would emit signals to passing smart phones that could then pull up a deal offering a discount on a shirt on a nearby rack.
LED bulbs are more costly than standard lightbulbs but have been dropping in price. Bytelight’s bulbs will be only “marginally” pricier than LEDs.
Ultimately, by using Bytelight’s technology, finding that last item in the grocery store will be infinitely easier, and you may even get a discount at checkout.
“We want people to think about light bulbs in an entirely new way,” Ganick told Technology Review.
Bytelight was spun off from the NSF Smart Lighting Engineering Research Center (ERC) and supported by Boston University’s Photonics Center.
-Samantha Gordon (COM ’12)
“BU Startup, Bytelight, Receives Funding From U-Launch”
Improving the efficiency of light emission to enable laser development from group-IV semiconductors, which provide the leading materials platform of microelectronics, has the potential to revolutionize photonics research and improve everything from on-chip data transmission to biochemical sensing and wireless optical communications.
At Boston University, Professor Roberto Paiella (ECE), Cicek Boztug (PhD ’14), and Faisal Sudradjat (PhD ’12) are collaborating with researchers from the University of Wisconsin-Madison to overcome challenges associated with the radiative properties of silicon, germanium and related alloys, all of which don’t emit light very efficiently due to a basic materials property – their so-called indirect energy bandgap.
They discovered that germanium nanomembranes (i.e., single-crystal sheets no more than a few tens of nanometers thick), when mechanically stressed, can be used to overcome this fundamental limitation and serve as great light emitters, particularly for the mid-infrared spectral region where many biological and chemical species of interest have distinctive absorption lines.
“There have been a lot of efforts to make silicon and germanium efficient photonic active materials,” Paiella said. “Our method has proven to be highly effective.”
The research team wrote a paper on their work titled, “Direct-Bandgap Light-Emitting Germanium in Tensilely Strained Nanomembranes,” and the Proceedings of the National Academies of Sciences of the United States of America recently published their findings.
“We were able to demonstrate that tensilely strained germanium is a good candidate for chip-level integration of electronics and photonics for mid-infrared applications,” said Boztug. “Potentially, this new development could lead to CMOS-compatible biochemical sensors as well as secure free-space communication devices integrated on silicon chips.”
Paiella said that using germanium nanomembranes to emit light is a unique idea in photonics research and that it could enable the development of silicon-compatible diode lasers, which represent the “missing link” for the full integration of electronic and photonic functionalities on the same materials platform.
“If you can make a laser this way, you can integrate laser sources directly on electronic chips,” said Paiella. Potential results could include improved on-chip data transfer and better optical sensing.
For Boztug, working on this project has been eye-opening. Not only did she learn about the mechanical properties of nanomembranes through UW-Madison’s Professor Max Lagally, she also had an opportunity to learn about photonics from for her advisor, Paiella.
“I feel very lucky to have an outstanding advisor like him and realized that I learn something new every time we meet,” she said.
Her research at BU centers around group-IV photonics, an area that she said offers great potential for information processing as well as biological and chemical sensing applications.
“I believe that our studies could lead to a new era for on-chip biochemical sensing applications by combining the well-established silicon microelectronics with our light-emitting germanium membranes,” she said.
-Rachel Harrington (email@example.com)