Rice-led project aims to boost performance on every chip

Wednesday, April 8, 2009

DARPA awards $16 million to Rice University to improve compilers

The Defense Advanced Research Projects Agency (DARPA), as part of its Architecture Aware Compiler Environment Program, has awarded Rice University $16 million to develop a new set of tools that can improve the performance of virtually any application running on any microprocessor.

The PACE project -- short for "platform-aware compilation environment" -- centers on ubiquitous computer programs called compilers. All microprocessors -- not just those in PCs but also the ones powering cell phones, game systems, cars and even electronic toys -- have their own compilers to translate human-written computer applications into the binary 1s and 0s that a processor can execute.

"To use a new computer system effectively, an applications programmer needs a high-quality compiler, one that can translate the application in a way that achieves a reasonable fraction of the available performance," said Keith Cooper, Rice’s John and Ann Doerr Professor in Computational Engineering and a principal investigator on the PACE project. "Unfortunately, it typically takes about five years to develop a high-quality compiler for a new computer system, and because that's longer than the effective life cycle of most microprocessors, we rarely see a case where applications make good use of a processor's resources."

The plethora of microprocessors only adds to the problem. Most electronic gadgets -- everything from iPhones to digital hearing aids and GPS systems to antilock brakes -- have a specialized "embedded" microprocessor. New personal computers and laptops typically contain two or more general-purpose processors on a "multicore" chip from Intel or AMD, as well as a high-performance graphics processor, a sound card processor and other specialized processors. Sony's PlayStation 3 game system has an IBM Cell Broadband Engine that contains one general-purpose microprocessor and eight specialized processors.

Cooper said the military's interest in funding PACE stems from its heavy reliance on computing, ranging from supercomputers for global weather forecasts to portable devices used by infantry.

"When a compiler translates human-written code into executable code, it makes myriad choices that have a direct impact on how fast the application runs, how much power it uses and how much memory it uses," Cooper said.

The tools PACE project researchers hope to build will cut the time needed to create high-quality compilers. In addition, the PACE team will learn as it goes, measuring and weighing the goals, capabilities and performance of each processor, to create compilers that are optimized for particular situations.

Krishna Palem, Rice's Ken and Audrey Kennedy Professor of Computer Science, said, "It is a rare treat to be working with this 'dream team' and continue Rice's rich tradition in compiler research. PACE involves many innovations using radical ideas intended to allow compilers to learn and adapt, much as humans do during infancy."

The PACE "dream team" includes researchers from Rice, Texas Instruments, ET International, Ohio State University and Stanford University. Rice's team consists of five pre-eminent compiler researchers: Keith Cooper, John Mellor-Crummey, Krishna Palem, Vivek Sarkar and Linda Torczon. Each will lead part of the research and development activity. Researchers outside Rice include Reid Tatge of Texas Instruments; Rishi Khan, director of research and development at ET International; P. Sadayappan, professor of computer science and engineering at Ohio State; and Sanjiva Lele, professor of aeronautics and astronautics and of mechanical engineering at Stanford.

Vivek Sarkar, Rice's E.D. Butcher Chair in Engineering and professor of computer science, likened PACE's challenge to the famous test computer scientist Alan Turing posed in 1950: A computer could only be said to be truly intelligent if its actions were indistinguishable from a human's.

"This is akin to a Turing Test for compilers," Sarkar said. "Our goal is to enable PACE tools to be used as a substitute for the time-consuming human expertise that is currently needed to improve the quality of compilers for any given platform.

"The challenge is daunting," he said. "It's not just hard, it is DARPA-hard."

Because the PACE project will focus on portable performance, Cooper said, the researchers will rely on vendor-supplied compilers -- for languages such as C and Fortran -- to perform the final steps of code generation for the target systems. The output of the PACE tools will be a distinct version of an application's code for each kind of processor in the system. Each of those codes will be specifically optimized for the processor, the surrounding system and the vendor compiler.

Sarkar said, "In this way, the PACE system will manage the application performance that can be achieved using less-ambitious compilers for the component processors."

HIV treatment test closer to manufacture with new $7.3 million grant

Tuesday, February 17, 2009

An initiative that is developing a rapid and inexpensive test to analyze the immune system of people living with HIV/AIDS has received a $7.3 million grant from the Bill & Melinda Gates Foundation, it is announced today. If successful, the test would improve healthcare workers' ability to determine the best treatment for their patients.

The CD4 Initiative at Imperial College London was established to develop an easy to use point-of-care test with a cost of around $2 that can rapidly measure the numbers of CD4+ T-cells in a person's blood, without using electronics or mechanical parts.

CD4+ T-cells are critical for a healthy functioning immune system and are slowly destroyed during the course of HIV infection. When the numbers of CD4+ T-cells in a person's blood drop, this makes them increasingly vulnerable to illness. Healthcare workers rely on a CD4 count when making decisions about how HIV-positive patients should be treated and when they should begin antiretroviral therapy. The new test would enable patients to find out within minutes if they should begin antiretroviral treatment.

Imperial's academic and industrial partners in the CD4 Initiative have worked since 2007 to devise the new test. Teams from Beckman Coulter, Inc (USA), Macfarlane Burnet Institute (Australia) and Zyomyx, Inc (USA) have already developed three prototypes, one of which will be chosen to be manufactured and mass produced in 2010.

The new test will work with a finger-prick blood sample and will have a simple read-out. One of the new prototypes has a design similar to that of a home pregnancy test.

The majority of patients in the developing world do not currently have access to CD4 testing because it is expensive and requires specially trained operators. Where testing facilities exist, it is often too difficult for people in rural areas to reach them. For those who are tested, it can take weeks to obtain results.

"There has been a lot of progress increasing access to life-saving HIV drugs in the developing world, but the lack of access to essential diagnostic tests like a CD4 test is a major barrier to providing the best possible care," said Dr Hans-Georg Batz, Director of the CD4 Initiative from the Division of Medicine at Imperial College London. "The majority of patients start anti-retroviral therapy based on symptoms alone. Research shows that if you wait until you're sick to start treatment, you have a much poorer outcome than if you start based on CD4 count. Our new test will have a huge positive impact for people living with HIV across the world."

Dr Steven Reid, the project manager for the CD4 Initiative at Imperial College London, added: "In resource-poor rural areas, patients sometimes have to walk miles to get to a clinic. Even if a traditional CD4 count is offered, the patients have to come back in a couple of weeks for the result. By then, some are too sick or cannot afford to return. For others, it may already be too late.

"The new test would eliminate the need to wait for treatment. As soon as a patient finds out that they are HIV positive, they could immediately find out if they need to start on antiretroviral therapy," added Dr Reid.

The CD4 Initiative is a multi-partner, international collaborative effort housed at Imperial College London. A 2005 grant of $8.6m from the Bill & Melinda Gates Foundation supported the first three years of the project. The $7.3 million grant announced today will support trials of the new test in developed and developing world countries, developing the manufacturing process, and initial production runs.

The CD4 Initiative is taking a project management approach to devising the new test. This approach, commonly used in industry, means that multiple research teams from academia, private companies and other institutions around the world work together collaboratively, using strict milestones and set timelines. The CD4 Initiative team believe their success to date demonstrates that this new kind of approach saves both resources and time and could be rolled out to other international health programs. This public/private sector collaboration is a successful example of this kind of management approach.

UQ research delves further into the mysteries of the brain

Monday, November 10, 2008

Researchers from UQ's Queensland Brain Institute will delve further into the mysteries of how our brains works thanks to more than $7.5 million in funding.

The grant, totaling $7,627,200, is part of National Health and Medical Research Council Program Grant funding announced today and is aimed at understanding how new cells are generated in the adult brain and how they produce functional changes.

Professor Pankaj Sah, one of four chief investigators, said the funding was a renewal of an existing project that had already unlocked many secrets of our most vital organ. Other chief investigators are QBI director Professor Perry Bartlett and Professors Seong-Seng Tan and Trevor Kilpatrick from the Howard Florey Institute.

“Our understanding of the brain has come a fair way in the past few years,” Professor Sah said.

“We have identified how new cells are born in our brain as adults and how they integrate into the neural circuitry.

“Our new focus is looking at behavioral outcomes and how those new cells become parts of our learning and memory functions.”

He described the work as trying to find out how new cogs can be added to an existing machine without “throwing a spanner in the works”.

“This will not only help us understand how the brain develops and how we learn and remember, but also how the brain recovers after disruptions such as stroke and dementias such as Alzheimer's disease affect those processes,” he said.

Professor David Siddle, Deputy Vice-Chancellor (Research), said the funding would take UQ research to another level.

“Understanding the brain and how it functions is one of the last frontiers of medical science,” Professor Siddle said.

“This research will bring us a step closer to learning what it is that makes this organ unique among the many in our amazingly complex bodies.

“I would also congratulate the other UQ researchers to receive funding in such a highly competitive grant environment.”

Professor Matt Brown, from UQ's Diamantina Institute, was another researcher to receive funding today, looking at the genes associated with osteoporosis, which could be used to develop predictive tests and new treatments

Professor Brown received $800,000 in an EU Collaborative Grant that will see him working with colleagues in Australia and Europe.

Five UQ researchers have also been awarded NHMRC Research and Practitioner Fellowships as part of the funding announcement: Professor David Craik, from the Institute for Molecular Bioscience; Associate Professor Joseph Lynch, from the Queensland Brain Institute; Associate Professor Alexander Khromykh, from the School of Molecular & Microbial Sciences; Associate Professor Frederic Meunier, from the Queensland Brain Institute; and Associate Professor Timothy Florin, from the School of Medicine.

CU-Boulder Awarded $7.2 Million for Liquid Crystal Research Center

Tuesday, September 16, 2008

The National Science Foundation has renewed its support of the University of Colorado at Boulder's Liquid Crystal Materials Research Center, awarding the university a $7.2 million grant.

This is the third round of multiyear funding the center has received through the NSF Materials Research Science and Engineering Centers program. The MRSEC program supports a network of collaborative university research centers that pursue basic materials research and develop and test materials for commercial and consumer applications.

Founded in 1995, the Liquid Crystal Materials Research Center brings together faculty, graduate and undergraduate students from the physics, chemistry and biochemistry, and chemical engineering departments to study and develop new liquid crystal materials and uses. CU-Boulder physics Professor Noel Clark, center director, and chemistry and biochemistry Professor Dave Walba, associate director, began working together on liquid crystals research in 1983.

Liquid crystals are organic materials related to soap. They behave both like a solid and a liquid and are highly sensitive to changes in temperature and voltage. This makes them ideal for use in information display applications such as computer and cell-phone screens, watch faces, calculators and flat-panel televisions, said Clark.

"Materials research has had great importance for the university for many years," said Stein Sture, vice chancellor for research and dean of the Graduate School. "There are graduate students who come here specifically for this program and then go on to successful careers around the country."

In addition to basic materials research, the center has been successful at developing new applications for liquid crystals, said Clark. The center has spun off six "daughter" companies, including two-year-old Naxellent in Broomfield, which develops liquid crystal technology for solar control panels.

The center also is collaborating with three other companies to develop non-display applications for liquid crystals that will improve the efficiency of solar heating, Clark said. For example, windows filled with a temperature-responsive liquid crystal might absorb light when it is cold out and turn opaque to reflect light when it's hot.

"To bring solar applications to the next level is a major challenge and will require extensive basic research and creativity," said Sture. "If they crack that nut, it will be a major contribution to the world."

In addition to the six-year, $7.2 million NSF grant, the state of Colorado and CU-Boulder will contribute $800,000 a year in matching funds to the center.

Clark called the grant "life" for the center. "It means we can pursue a lot of new ideas," he said. "We've been pretty good at staying ahead of the field, and we hope to keep on doing that.

Researcher Awarded Grant to Develop Technology for Generating Insulin-Producing Cells to Treat Diabetes

Tuesday, September 16, 2008

Kaiming Ye, assistant professor of biomedical engineering at the University of Arkansas, has received a $244,000 grant from the National Science Foundation to develop a new technology that will generate glucose-responsive, insulin-producing cells from human embryonic stem cells. The research will lead to a new generation of cell-based therapies for treating diabetes.

"The outcome of these studies, made possible by this award, will ultimately advance stem-cell and tissue engineering,” Ye said. “Specifically, our technique will offer renewable cell sources for transplanting tissue that restores near physiological insulin-secretion capacity in people with both type-1 and type-2 diabetes.”

As Ye mentioned, stem-cell and tissue engineering hold great promise as an effective therapy for diabetes. However, before glucose-responsive, insulin-expressing cells can be regenerated and transplanted to diabetes patients, scientists must gain a better understanding of the origin of the human pancreas during embryonic development. The pancreas produces various proteins and enzymes that, among many functions, regulate blood-sugar levels.

Ye’s work will strengthen this understanding. Specifically, the research team will build three-dimensional cultures of pancreatic stem cells. Previous studies have been limited to two-dimensional cell cultures that fail to mimic three-dimensional environments of the human body.

“To the best of our knowledge, 3-D pancreatic differentiation of human embryonic stem cells has not been reported,” Ye said. “Most of our knowledge about the origin and embryonic development of the human pancreas is gained from studies of mice and cell differentiation in two-dimensional culture systems. However, a significant body of evidence suggests that 2-D cultures fail to imitate the in vivo environment where tissues are organized in a three-dimensional architecture. Thus, knowledge acquired from 2-D human embryonic cell differentiation does not correctly reflect what happens in the body.”

Ye’s research team conceived the idea of three-dimensional cultures based on their work on mouse embryonic stem cells in which they used a simple collagen scaffold to promote the maturation of insulin-secreting beta cells derived from the embryonic stem cells of mice. They observed a significant improvement of the maturity of beta cells differentiated from mouse embryonic stem cells. These experiments encouraged the researchers to think about the possibility of elevating the maturity of beta cells with human embryonic stem cells.

Specifically, the project will:

• Establish a three-dimensional collagen scaffolding of human embryonic stem cell cultures and determine the proliferation, differentiation and adhesion of cells within the cultures.

• Characterize the differentiation of human embryonic stem cells into mature pancreatic beta cells, which produce insulin, in three-dimensional cultures through profiling of gene expression.

• Construct fluorescent markers for tracing the differentiation of human embryonic stem cells into pancreatic beta cells in three-dimensional cultures.

• Demonstrate the generation of glucose-responsive, insulin-producing cells from human embryonic stem cells in three-dimensional cultures.

Initial studies will focus on in vitro experiments. Subsequent studies will test in vivo functions of pancreatic cells in diabetic animal models through tissue transplantation.

Nanophotonics Research Links UCSD, Sun Microsystems and the Future of Computing

Thursday, March 27, 2008

Electrical engineers at UC San Diego together with researchers from Sun Microsystems and Stanford University will receive $44 million from DARPA (The Defense Advanced Research Projects Agency) to develop connections between computer chips using light rather than wires. Connecting hundreds or thousands of chips within supercomputers via optical links capable of carrying tens of billions of bits of data per second would lead to faster, more energy-efficient and compact computers.

This vision for the future of computing is grounded in the field of nanophotonics – an area of particular strength at UCSD’s Jacobs School of Engineering. In the early 1990s, electrical engineering professor Shaya Fainman began working with light on the sub-wavelength scale and helped to develop what is now called the field of nanophotonics.

“We are working on CMOS compatible nanophotonic devices that are manufactured with standard lithographic tools,” said Fainman, who was recently named the Cymer Inc. Endowed Chair in Advanced Optical Technologies.

UCSD will receive about half a million dollars from DARPA for this UCSD-industry project, and Fainman noted that this award highlights the well established and fruitful collaboration between UCSD’s Jacobs School of Engineering and Sun Microsystems.

“This partnership is an example of how the Electrical Engineering Department is demonstrating continued leadership within the field of optical communications and pushing the limits in areas of fundamental importance to the future scaling of microelectronic systems. Our long term investment in the field of photonics has reaped huge rewards for the Jacobs School of Engineering,” said Larry Larson, Chair and Professor, Department of Electrical and Computer Engineering.

In a paper in the July 2007 issue of Physical Review Letters, Fainman and coauthors from the Jacobs School and Sun Microsystems describe their “free space optics on a chip” configuration. Such a configuration would allow light to propagate freely in the slab of silicon, while interacting with discrete optical components that are located along the propagation direction.

“We believe that this new concept may become essential for applications such as optical interconnections, information processing, spectroscopy and sensing on a chip,” Fainman and colleagues write in their Physical Review Letters paper.

The work chronicled in the Physical Review Letters paper is tied to Sun Microsystem’s vision for a “macrochip” system in which ultranarrow silicon channels called waveguides shuttle beams of light from chip to chip – thus bypassing the wires that currently act as a major bottleneck for today’s computer designers.

“Optical communications could be a truly game-changing technology—an elegant way to continue impressive performance gains while completely changing the economics of large-scale silicon production,” Greg Papadopoulos, chief technology officer and executive vice president of research and development for Sun Microsystems said in a statement. Papadopoulos is an alumnus of UCSD’s Jacobs School of Engineering.

A host of other researchers at the Jacobs School are working in the area of photonics and optical communications. Additional details are available by searching the Jacobs School faculty database using related keywords.

Grant Focuses on Link Between Chemical Exposure and Prostate Cancer

Wednesday, March 26, 2008

With questions lingering about the estrogen-mimicking chemical Bisphenol A, a University of Cincinnati (UC) environmental health expert hopes to shed new light on the relationship between the man-made chemical and prostate cancer.

Shuk-mei Ho, PhD, chair of UC’s environmental health department, and her University of Chicago colleague have received nearly $2.6 million to study the mechanism by which Bisphenol A (BPA) exposure in the womb or in infancy may affect prostate cancer later in life.

Ho and co-principal investigator Gail Prins, PhD, reported the first evidence of a direct link between chemical exposure while in the womb and prostate cancer development later in life in the June 2006 issue of the journal Cancer Research.

In that laboratory study, the duo found that animals exposed to low doses of the natural human estrogen estradiol or the environmental estrogen BPA during fetal development were more likely to develop an early form of prostate cancer in humans (prostatic intraepithelial neoplasia) than those who were not exposed.

Their findings suggested that exposure to environmental and natural estrogens during fetal development could affect the way prostate genes behave, leading to higher rates of prostate disease during aging.

BPA is a chemical regularly used in the manufacture of plastics that can leach out when heated. It is one of many man-made chemicals known as “endocrine disruptors,” which permanently alter the function of the endocrine system by mimicking the role of the body’s natural hormones. Hormones are secreted through endocrine glands and dispersed to serve different functions throughout the body.

Ho and Prins will use this new five-year grant from the National Institute of Environmental Health Sciences to assess whether genes that are epigenetically regulated by estrogen exposures during fetal development play a direct role in prostate cancer development later in life.

The researchers believe that this increased sensitization to prostate cancer is an “epigenetic” effect of exposure to estrogen earlier in life. Epigenetics is an emerging field targeting heritable changes in gene expression that do not cause mutations in DNA.

“We know developmental exposure to natural or environmental estrogens may predispose an individual to prostate cancer with age, but the molecular underpinnings of this phenomenon are unclear,” explains Ho.

To explore their hypothesis, Ho and Prins will use two laboratory models designed to reflect the characteristics of a developing male: one in an animal and the other using human prostate-like structures. Their goal is to identify genes that undergo changes when exposed to low-dose environmental and natural estrogens and establish a dose-response relationship between prostatic BPA and the developmental window of susceptibility.

“We need to understand how these neonatal estrogens imprint or transform the prostate gland,” says Ho. “We know now that they are regulated, in part, by epigenetic mechanisms. Now we must determine if those epigenetic factors play a direct role in predisposing an individual to cancer later.”

According to the U.S. Centers for Disease Control and Prevention, more than 95 percent of Americans age 6 or older have measurable BPA in their blood—and higher concentrations of the chemical have been found in amniotic fluid, placenta and fetal tissues.

Honeybee Researcher To Unravel Properties Governing Lifespan With Support From Norway

Monday, March 24, 2008

Gro Amdam, associate professor in the School of Life Sciences at Arizona State University, has been awarded two grants totaling the U.S. equivalent of about $1.4 million from the Norwegian Research Council to investigate biochemical factors and social life history properties that can influence aging and longevity in honeybees. Amdam also is with the Department of Chemistry, Biotechnology and Food Science at the Norwegian University of Life Sciences in Norway.

The first study will focus on the molecular properties of honeybee vitellogenin, a protein which, Amdam says, acts at the intersection between social behavior and aging. The second project, to be headed by Amdam’s postdoctoral fellow Siri-Christine Seehuus in Norway, will examine the genetic and endocrine factors, which may determine longevity in diutinus workers, a specialized sub-caste of honeybees.

In a series of previous studies, Amdam has shown that vitellogenin protein affects aging rate and endocrine signaling in honeybees. In addition, separate studies conducted with Robert Page, director of ASU’s School of Life Sciences in the College of Liberal Arts and Sciences, demonstrated that the protein also influences social behavior, longevity and sensory responsiveness.

“Generally, vitellogenin is described as a conserved yolk protein found across a broad range of egg-laying species. The functions of proteins homologous to honeybee vitellogenin therefore have been studied primarily in the context of female reproduction,” Amdam says. “New data from my laboratory suggests, however, that the protein can be active in signal transduction”

Amdam hopes to understand more about the structural and binding properties for the honeybee vitellogenin protein through examination of synthesized protein fragments, combined with crystallography and spectroscopy. Her intention is to unlock how the protein can have pleiotropic effects on honey bee social organization which also may open a window onto mechanisms that enabled honey bee social life to emerge.

Amdam and Seehuus will both exploit the plasticity found in honeybee social life history in their work examining the causal basis of the extreme longevity of honeybee diutinus workers (up to 1 year, in contrast to the normal lifespan of about 2 months). While Seehuus will focus on endocrine regulation, Amdam will study the role of a key social factor, the presence versus the absence of young brood.

“Since sister honeybees can be both short-lived and extremely long-lived, it is clear that diutinus development within a colony is not determined by genetic predisposition,” Amdam notes. “Rather, diutinus bees develop as a function of social change when the young brood (honeybee larvae) is removed from the nest. Preliminary results from my lab in Norway point to a major effect on lifespan of pheromones released by the brood.” Amdam graduate students have found that exposing the workers to brood phermones alone (using synthetics in the absence of actual brood) prompt the diutinus workers to build up particularly large body reserves of proteins and fats which likely have positive effects on survival.

Next, Amdam will study how the dual effects of brood, that is the physiological load of nursing the larvae and exposure to brood pheromones, translate into levels of individual gene- and protein expression, storage dynamics of tissues, and at the level of behavior, food intake and feeding.

Amdam expects that together the planned studies will unravel patterns of interplay, from molecular- to social mechanisms that can govern lifespan in social species.

Amdam has made key discoveries in the genetic, physiological and behavioral mechanisms underlying division of labor, caste development, and advanced understanding around the evolution of social life strategies, including aging, in social insects. Since 2006, her work, primarily using the honeybee as a model organism, has been published in professional journals as varied as Nature, Science, Experimental Gerontology, Proceedings of the National Academy of Science (PNAS), Behavioural Brain Research, Public Library of Science (PLoS), American Naturalist and Advances in Cancer Research. In 2007, Pew Charitable Trusts selected Amdam to be a Pew Scholar in biomedical sciences and she had the distinction of being named “Outstanding Young Investigator” by the Research Council of Norway.