Today at Berkeley Lab

CRD’s David Patterson to Help Lead ‘Crucial’ Google Project

Patterson, who recently retired as a UC Berkeley professor, is now a key part of the team behind a critical chip that Google uses for artificial intelligence processing. Without the chip, it is estimated Google would have to double its data centers to support even a limited amount of voice processing. More>

You can leave a comment. View the Comment Policy.

Matthias Vallentin Takes a “VAST” Step Forward in Cyber Security

Postdoctoral researcher Matthias Vallentin of the Computational Research Division (CRD) is working with UC Berkeley computer science professor and CRD network researcher Vern Paxson to develop VAST, a system that will help forensic security analysts pinpoint how much of an organization’s computer network has been compromised, and where. More>

You can leave a comment. View the Comment Policy.

CRD’s Scott Beamer Wins Distinguished Dissertation Award

Scott Beamer, a postdoc in the Computational Research Division’s Computer Architecture Group, will receive the 2016 SPEC Kaivalya Dixit Distinguished Dissertation Award at the 8th ACM/SPEC International Conference on Performance Engineering being held April 22-27 in L’Aquila, Italy. More>

You can leave a comment. View the Comment Policy.

Recognition Software Drives Matches Across Multiple Science Domains

A team led by Daniela Ushizima of the Computational Research Division has built a Python-based tool for content-based image retrieval (CBIR) capable of searching relevant items from large datasets, given unseen samples. Named “pyCBIR,” the tool can be used to catalog and retrieve images from different science domains, such as biology, materials research, and geology. More>

You can leave a comment. View the Comment Policy.

Rare Supernova Discovery Ushers in New Era for Cosmology

Using an automated supernova-hunting pipeline based at NERSC, astronomers have captured multiple images of a gravitationally lensed Type Ia supernova. This detection is currently the only one of its kind. Lab scientists have a method for identifying more of these events using existing wide-field surveys. Check out yesterday’s Facebook Live Q&A on this research. More>

You can leave a comment. View the Comment Policy.

Watch Berkeley Lab’s First ‘Facebook Live’ Today at 1 p.m.

Join astrophysicist Peter Nugent of the Computational Research Division and UC Berkeley graduate student Danny Goldstein in our first Facebook Live event today at 1 p.m. They’ll take questions and discuss their latest supernova research, to be announced today at 11 a.m. on the Lab’s News Center.

You can leave a comment. View the Comment Policy.

Daniela Ushizima Unveils New Science Search Tool for the Web

Googling a photo of an obscure band may be a snap, but not so for scientific images. That’s why the Computational Research Division’s Daniela Ushizima decided to create a tool tailored to scientific datasets that are in databases not immediately obvious to the web. More>

You can leave a comment. View the Comment Policy.

Researchers Catch Extreme Waves with Higher Resolution Modeling

Resolution matters, particularly when it comes to producing accurate models of hurricanes and the extreme waves they generate. A new Berkeley Lab study shows that high-resolution simulations captured tropical cyclones and waves that low-resolution ones missed. Better extreme wave forecasts are important for coastal cities, the military, and the shipping industry. More>

You can leave a comment. View the Comment Policy.

CRD’s Julian Borrill Receives NASA’s Exceptional Public Achievement Medal

Borrill was honored for “conceiving and implementing the mission-critical high performance computing system for Planck data analysis.” The Planck satellite mission is a joint project between the European Space Agency and NASA. Borrill has served as the U.S. team’s computational systems architect for more than 10 years. More>

You can leave a comment. View the Comment Policy.

New Technique Produces SHARP 3-D Images

A revolutionary X-ray imaging technique known as SHARP (Scalable Heterogeneous Adaptive Real-time Ptychography) combines diffraction and microscopy with applied mathematics and supercomputing to quickly turn high-throughput datasets into the sharpest 3-D images ever produced. More>

You can leave a comment. View the Comment Policy.