Distinguished Lecture Series

OMI: The Invisible Technology that will Revolutionize Supercomputing and AI

Prof. Harm Peter Hofstee
Delft University of Technology
Distinguished Research Staff, IBM Austin Research Laboratory

3:00pm Thursday 14 February, 2019, ITE325, UMBC

In this talk, we present some major trends in compute, memory/storage, and networking, and for each we will discuss how OpenCAPI Memory Interface (OMI) and related interface technologies are set to transform how we build, program, and think about our computer systems. For the first of these trends, it allows us to compensate for the reduced growth in processor performance (per dollar) and performance per Watt. Accelerators are sharing memory and other resources over NVLink or OpenCAPI with conventional IBM POWER cores and are driving performance in the world’s largest supercomputers and IBM’s systems are targeting AI and other workloads. The second addresses the reduced improvement in memory cost and capacity. OMI allows us to use technologies other than DRAM as memory, and because many of these technologies are nonvolatile, the line between memory and storage is becoming blurred. The third, OpenCAPI-based networking leverages the rapid improvements in cost per Gb/s and allows us to contemplate systems that extend memory beyond the node using commodity infrastructure.

Harm Peter Hofstee is a Dutch physicist and computer scientist who currently is a distinguished research staff member at the IBM Austin Research Laboratory, USA, and a part-time Professor in Big Data Systems at Delft University of Technology, Netherlands. Hofstee is best known for his contributions to Heterogeneous computing as the chief architect of the Synergistic Processor Elements in the Cell Broadband Engine processor used in the Sony PlayStation 3, and the first supercomputer to reach sustained Petaflop operation. His early research work on coherently attached reconfigurable acceleration on POWER7 paved the way for the new coherent attach processor interface on POWER8. Hofstee is an IBM Master Inventor with more than 100 issued patents and a member of the IBM Academy of Technology. Hofstee was born in Groningen and obtained his master’s degree in theoretical physics of the University of Groningen in 1988. He continued to study at the California Institute of Technology where he wrote a master’s thesis Constructing Some Distributed Programs in 1991 and obtained a Ph.D. with a thesis titled Synchronizing Processes in 1995. He joined Caltech as a lecturer for two years and moved to IBM in the Austin, Texas, Research Laboratory, where he had staff member, senior technical staff member and distinguished engineer positions.

The post talk: OMI, Invisible Technology that will Revolutionize Supercomputing and AI; 3pm Thr Feb 14, ITE325 appeared first on Department of Computer Science and Electrical Engineering.

The Shriver Center still has some opportunities available for UMBC students interested in Service-Learning & Community Engagement for the spring 2019 semester:

The Summer 2019 Orientation season is approaching and the Office for Academic and Pre-Professional Advising is looking for new members to join its team of orientation advisors! 

 Faculty, staff and graduate students are invited to apply. 

Graduate students and 9½-month faculty members will receive up to $100.00 per session.

12-month faculty members' compensations will be sent to their departments.

Staff members are not eligible to receive additional compensation, but they will have an opportunity to improve their advising competencies, work with a diverse body of students, and enhance their leadership and interpersonal skills.  

Comprehensive training will be provided for all orientation advisors on June 4 and 5.

Want to learn more about this opportunity? Attend an info session:

February 27, 2019    12:00pm-12:45pm   SHER 220 (B-WING)

March 13, 2019         12:00pm-12:45pm   SHER 220 (B-WING)

March 27, 2019         12:00pm-12:45pm   SHER 220 (B-WING)

April 10, 2019            12:00pm-12:45pm   SHER 220 (B-WING)

April 24, 2019            12:00pm-12:45pm   SHER220 (B-WING) 

We are currently accepting new applications through May 1st.

Contact Monique Ransom, Asst. Director, OAPA at mransom@umbc.edu or (410)455-2451 with questions.

Hello ISCOMers,

We begin our Cybersecurity month TODAY with Dr. Forno who will discuss the foundations of modern cyber defense, the current threats that governments and independent organizations face, and career prospects within the world of cybersecurity. Find us in ITE 241 from 12:00-12:50pmTODAY and don't miss out!

Hope to see you all there!

This story was written by Sarah Hansen and first appeared on news.umbc.edu

“The drug you design ten years from now may already be obsolete,” Ivan Erill says. In a new study in Frontiers in Microbiology, Erill and colleagues describe how bacteria that existed hundreds of millions of years ago were already resistant to an antibacterial drug not invented until the 1930s. Once farmers began using the new class of drugs in agriculture, resistance spread quickly. As the bacteria were exposed to the drugs on a large scale in soils and waterways, antibiotic-resistant strains began appearing in hospitals within a decade.

How is this all possible? Antibiotic resistance has much deeper roots than most people realize, Erill explains. Many antibiotics used to treat bacterial infections today are based on molecules bacteria naturally produce to out-compete their neighbors. “Your competitors are not just going to stand by,” says Erill, professor of biological sciences, “so over millions of years they are going to develop resistance. That’s a given.”

But the drugs featured in this research are not natural antibiotics. They are synthetic compounds produced by humans. “With synthetic drugs, that’s a different picture altogether,” Erill says. “It’s not a given that you would find resistance.” That’s why the research team was surprised when they did—hundreds of millions of years before the drugs were invented.

Tracking down the source

The synthetic antibacterial drugs in question, called sulfonamides, target an enzyme involved in a pathway necessary for DNA synthesis. If an organism can’t replicate its DNA, it can’t reproduce, so its population quickly dwindles to nothing. The bacteria that are resistant to sulfonamides have modified genes for the target enzyme, called sul genes (for sulfonamide resistance). These genes allow the bacteria to continue reproducing in the presence of the drug, which means the antibiotic won’t work on them.

Most sulfonamide-resistant bacteria have sul genes in what are called mobile elements—small segments of DNA that can easily jump from one individual or species to another. Erill and colleagues used computational tools to figure out which bacteria species had sulfonamide-resistance genes in their chromosomes instead of in mobile elements, indicating that they were the original sources of the resistance. They showed that two groups of bacteria harbor chromosomal sul-like genes.  

Ivan Erill. Photo by Marlayna Demond ‘11 for UMBC.

“You can use algorithms to reconstruct the most likely evolutionary history that explains the development of sulfonamide resistance,” Erill says. Those algorithms allowed the researchers to confirm that the resistance existed 500 to 600 million years ago. To further verify their results, the researchers inserted copies of the chromosomal sul-like genes into bacterial cells in the lab. When exposed to sulfonamides, the cells grew just fine, confirming that the genes confer resistance.

Pure chance

The question remains, though: “How can you explain that bacteria 500 million years ago were resistant to a substance that didn’t yet exist?” asks Erill. While there is an small chance some organism was producing a sulfonamide-like compound hundreds of millions of years ago and resistance evolved in response to that pressure, Erill and his team put forth a different argument.

“There is an enormous amount of bacterial genetic diversity,” Erill says. Miquel Sanchez, a Ph.D. student at the Universitat Autònoma de Barcelona and the first author on the paper, adds, “Resistant variants of the antibacterial target could be present in the global genetic pool even before microbes are exposed to them.”

So, the reason these two bacterial groups were resistant to a compound that had never existed? Erill says, “We argue that this is just pure chance.”

Rogue bacteria

This research has big implications for the development and use of future antibiotics. If scientists develop a new antibiotic, “it is well possible that there might be one bacterium in the world that is already resistant,” Erill says. With conservative use of the antibiotic in human patients, though, it’s unlikely that particular bacterium would ever be exposed to the antibiotic and spread its resistance.

But, “if you overuse the drug, especially in an agricultural setting, where the drug slowly permeates into the soil, waterways, and underground water reservoirs,” you’re exposing “this huge population of bacteria that otherwise would never be bothered by antibiotics or synthetic drugs,” Erill says. And, based on the team’s analysis, if the one bacterium that is already resistant is exposed to the drug, “this variant that is resistant will jump to other species in a matter of years.”

“For me, it’s especially a warning against using antibiotics in farm settings,” Erill says. New drugs are typically tested using disease-causing bacterial species, “but maybe that’s not enough,” he says. “Maybe you should do broader testing, especially on the non-usual suspects, like soil bacteria.”

Erill also says the new finding points toward using combination therapies more often. A bacterium in nature might harbor a chance resistance to one compound it’s never encountered, but it’s unlikely to be resistant to two. If doctors use two drugs at once, it is likely that one of the drugs will kill the bacteria, preventing it from spreading its resistance to the other drug.

These new research findings could affect how well superbugs are kept at bay and the effectiveness of new antibiotic treatments. Whether the agricultural industry and drug developers heed the team’s advice remains to be seen.

Image: From left to right, the authors of the paper: Ivan Erill, Pilar Cortés, Jordi Barbé, and Miquel Sánchez-Osuna. Photo by Ángela Martínez Mateos.

This story was written by Sarah Hansen and first appeared on news.umbc.edu

The potential spread of disease from farmed oysters to wild oysters is a frequent concern for oyster producers and consumers alike. Contrary to common perceptions, new UMBC research in Aquaculture Environment Interactions has found that properly managed oyster aquaculture operations can actually help limit the spread of disease among wild oysters.

“This is another line of evidence saying that oyster aquaculture can be a good thing,” says Colleen Burge, assistant professor of marine biotechnology at UMBC, and a co-author on the study.

“The established way of thinking is that disease spreads from aquaculture, but in fact aquaculture may limit disease in nearby wild populations,” adds Tal Ben-Horin, a postdoctoral fellow at the University of Rhode Island and the lead author on the study.

Ph.D. student Sandy Picot works in Colleen Burge’s lab.

Creative conversations

At a workshop funded by the NSF as part of the Ecology of Infectious Marine Disease Research Coordination Network, Burge co-led a session on disease transmission among sea creatures. Prior to the workshop, Ben-Horin began working on mathematical models to examine the interactions between farmed oysters, wild oysters, and the common oyster disease Dermo.

Once at the workshop, Ben-Horin, Burge, and others further explored how best to refine Ben-Horin’s models. “It was exciting to be in the room, talking to experts, bouncing around ideas about different biological systems,” Burge says.

Burge is accustomed to working with researchers from different backgrounds, because her lab is located at the Institute of Marine and Environmental Technology, a facility on Baltimore’s Inner Harbor that houses researchers from UMBC, University of Maryland Center for Environmental Science, and University of Maryland, Baltimore.

One of the results of the NSF-sponsored workshop was her new paper. Burge and colleagues from Rutgers University, the U.S. Department of Agriculture, and the Virginia Institute of Marine Science contributed their expertise regarding oyster biology and the ecology of the oysters’ environment, and Ben-Horin used that input to inform the model.

An Eastern oyster undergoes testing in Colleen Burge’s lab.

Oyster aquaculture boom a boon?

According to Burge, disease is a major factor in limiting wild oyster populations. Dermo, one of the most prevalent oyster diseases, is caused by a single-celled parasite that occurs naturally in the environment and proliferates in the tissue of host oysters. When infected oysters die, their decaying tissue spreads the parasite to other oysters. But it takes two to three years for the parasite to kill the oysters, and as long as the oysters are harvested before that time, the the spread of the parasite (which is harmless to humans) is limited.

This new research suggests that it’s time to “start thinking about oyster aquaculture as being a sink for disease, rather than a source,” says Burge. She adds that the study “furthers the idea that you can plant a large number of oysters in aquaculture, and it won’t (from a Dermo disease perspective) negatively affect the wild oysters if you harvest them early enough.”

That’s important, because “oyster aquaculture is growing in Maryland right now,” Burge says. In addition, conservation groups are using oyster reef restoration as a tool to improve the Chesapeake Bay’s water quality and ecosystem health overall.

“This paper suggests this growing aquaculture industry could actually support restoration efforts and also the wild fisheries,” Burge says. That is, as long as aquaculturists employ proper management techniques. Oyster farms that grow their product on the bottom surface of the Chesapeake Bay or other bodies of water are unlikely to recover all of their oysters, increasing rather than reducing the spread of the disease, explains Ben-Horin. Growing the oysters in cages or bags above the bottom solves that problem.

Colleen Burge at work in her laboratory.

It’s all connected

Beyond this new study, Burge’s research more broadly explores how oysters and other bivalves work as some of nature’s most efficient water filters. She examines how they filter out pathogens that affect a range of species. In addition to supporting the current work led by Ben-Horin, she recently received funding to investigate how oysters filter out an eelgrass pathogen from water.

Eelgrass populations have been on the rise in the Bay in recent years, which is a good sign for overall Bay health. “Eelgrass is a foundational species and can provide structure for lots of different animals to live in, including fish, scallops, and other invertebrates,” Burge says. In addition, it’s been shown to moderate the water’s acidity in its immediate vicinity, possibly counteracting the negative effects of ocean acidification.

Eelgrass might even affect oysters and their diseases. “Sometimes ecosystem-level effects are overlooked, but,” Ben-Horin says regarding the new research, “in this case they’re front and center.” He reflects, “Everything that happens in the water is connected.”

All photos by Marlayna Demond ’11 for UMBC.