New computer program automates chip debugging
Fixing design bugs and wrong wire connections in computer chips after they’ve been fabricated in silicon is a tedious, trial-and-error process that often costs companies millions of dollars and months of time-to-market.
Engineering researchers at U-M say it doesn’t have to be that way. They’ve developed a new technology to automate “post-silicon debugging.”
“Today’s silicon technology has reached such levels of small scale fabrication and of sheer complexity that it is almost impossible to produce computer chips that work correctly under all scenarios,” says Valeria Bertacco, assistant professor of electrical engineering and computer science and co-investigator in the new technology. “Almost all manufacturers must produce several prototypes of a given design before they attain a working chip.”
FogClear, as the new method is called, uses puzzle-solving search algorithms to diagnose problems early on and automatically adjust the blueprint for the chip. It reduces parts of the process from days to hours.
“Practically all complicated chips have bugs and finding them all is intractable,” says Igor Markov, associate professor of electrical engineering and computer science and another of FogClear’s developers. “It’s a paradox. Today, manufacturers are producing chips that are good for almost all applications, from e-mail to chess, but they can’t be tested for every possible condition. It’s physically impossible.”
For modern chips, the process of making sure a chip is free of bugs takes as much time as production.
“Bugs found post-silicon are often very difficult to diagnose and repair because it is difficult to monitor and control the signals that are buried inside a silicon die, or chip. Up until now engineers have handled post-silicon debugging more as an art than a science,” says Kai-Hui Chang, who received a doctorate in electrical engineering and computer science in August. He will present a paper “Automating Post-Silicon Debugging and Repair” at the International Conference on Computer-Aided Design Nov. 6 in San Jose, Calif.
Millions of Americans over 70 suffer from dementia
One in seven Americans over the age of 70 suffers from dementia, according to the first known nationally representative, population-based study to include men and women from all regions of the country.
About 3.4 million people, or 13.9 percent of the population age 71 and older, have some form of dementia, the study finds. The prevalence of dementia increased dramatically with age, from 5 percent of those aged 71 to 79 to 37.4 percent of those age 90 and older.
About 2.4 million of those with dementia, or 9.7 percent of the population age 71 and older, were found to have Alzheimer’s disease, the most common cause of dementia.
Published in the November 2007 issue of Neuroepidemiology, the study is based on data from 856 men and women who participated in the Aging, Demographics and Memory Study conducted in 2002 by researchers at the Institute for Social Research (ISR) and Duke University Medical Center and funded by the National Institute on Aging. The study was part of the larger ISR Health and Retirement Study, a nationally representative survey of Americans age 51 and older.
The study provides the first prevalence estimates based on a nationally representative sample of older adults from all regions of the United States, allowing policymakers and care providers to make better plans for caring for those with dementia and their families. Co-author Kenneth Langa, who is affiliated with ISR and with the U-M and Veterans Affairs healthcare systems, says the findings will assist in assessments of the impact of future treatment advances as they become available, and in gauging how well the country is doing in controlling and treating Alzheimer’s disease and other forms of dementia.
“As the elderly population in the United States grows, the number of individuals with dementia will certainly increase tremendously,” says economist David Weir, who directs the ISR Health and Retirement Study. “Planning for the long-term care needs of this vulnerable population will become increasingly important. This new data, used along with related data from the Health and Retirement Study, should increase our nation’s ability to successfully meet the needs of an aging U.S. population and those who love them.”
—Diane Swanbrow, News Service
Keys to business growth inspire little confidence
While business leaders recognize the importance of hiring the right people, providing adequate training, promoting teamwork and maintaining a positive organizational culture, few of them are confident that such initiatives actually can help drive growth of their companies, says a U-M business professor.
In her Leadership Pulse survey of top business executives, Theresa Welbourne, an adjunct professor of executive education at the Stephen M. Ross School of Business, asked more than 300 respondents in 18 industries about future growth opportunities at their firms.
She found that 26 percent of executives believe their human resource department or those responsible for “overall people and culture management” currently are key drivers of growth. Moreover, just 55 percent expect HR functions to contribute to growth in the next year, even though three out of four executives say that such issues can jump-start company growth — a conundrum she refers to as the “people paradox.”
“Rightly or wrongly, in many cases the HR function continues to be perceived as a paper-pushing function that is not directly impacting the bottom line, or in our case, not seeking out growth opportunities and pursuing them with line management,” says Welbourne, president and CEO of eePulse Inc. “This simply has got to stop. It is not acceptable that so much growth potential exists that is going untapped.”
Welbourne’s study asked business leaders to rate the degree of confidence they currently have and over the next year that these areas will drive growth: people and culture management; sales and support; manufacturing and delivery; finance and accounting; technology, research and development; and customer service.
When it comes to current performance, none of the six areas exceeded 40 percent. In other words, most business leaders have little confidence that any of these functions within their companies are presently leading the way in business growth.
“With this type of data, companies could engage in healthy dialogue about driving business growth and then develop a strong set of solutions and actions,” Welbourne says. “Data should be regularly collected, analyzed and used to supplement other data — sales, financial, production, etc. — in order to continually understand the drivers of growth and productivity.”
—Bernie DeGroat, News Service
Study examines veterans, factors that contribute to suicide
The largest and most up-to-date study of suicides among depressed veterans provides important new data that may help guide screening and treatment for all veterans.
The study finds that the predictors of suicide among veterans in depression treatment differ from those seen in the general American population, with younger, white, non-Hispanic men having the highest risk among the veterans.
The study, conducted by researchers from the VA Ann Arbor Healthcare System and the U-M Health System and Depression Center, will appear in the December American Journal of Public Health issue focused on veterans’ issues.
Veterans with substance abuse issues, and those who had been hospitalized for psychiatric reasons in the year before their depression diagnosis, also had a higher suicide risk. Surprisingly, older veterans who had been diagnosed with post-traumatic stress disorder in addition to depression had a lower overall rate of suicide than those without a PTSD diagnosis, perhaps because they were more likely to receive care through Veterans Affairs PTSD programs.
Though the study did not directly compare populations of veterans and non-veterans receiving treatment for depression, it does confirm that suicide rates were very high among depressed VA patients during the research period of 1999-2004, reinforcing the need for the VA’s recent initiatives to prevent suicide.
“Doctors learn about patient characteristics that might increase risk of suicide,” says first author Kara Zivin, a VA investigator and assistant professor in the Department of Psychiatry. “Typically, these are older age, male gender, and white race, as well as depression, and medical or substance abuse issues. But our study indicates that among veterans in depression treatment, the predictors of suicide may not be the same. We hope our findings will help guide physicians in understanding suicide risk among currently depressed veterans.”
In addition to Zivin and senior author Dr. Marcia Valenstein, an associate professor of psychiatry and leader of this study, additional authors are Myra Kim, John McCarthy, Karen Austin, Katherine Hoggatt and Heather Walters, all of the VA, Ann Arbor, the Medical School or the School of Public Health. Zivin, Valenstein and McCarthy are members of the Depression Center.
— Kara Gavin, UMHS Public Relations
