Skip to content

ASU’s Sol ranks amongst top-performing supercomputers globally

Brighter than NYU Langone Well being’s UltraViolet and packing extra firepower than Harvard College’s Cannon, Arizona State College’s Sol supercomputer has been listed among the many top-performing supercomputers globally.

The Top500, a global benchmarking group, has ranked Sol above Harvard, NYU and Johns Hopkins. As well as, IO500, one other benchmarking group, listed Sol among the many top-performing analysis and 10-node analysis supercomputers on the planet, together with the College of Cambridge, Stanford College and Georgia Tech. Each rankings are the very best ever for the college and the primary time Sol was ranked since its launch final yr. The Sol supercomputer at Iron Mountain Knowledge Heart. Photograph by Andy DeLisle Obtain Full Picture

“These rankings underscore the assist and empowerment Sol gives to ASU researchers,” says Sally C. Morton, government vice chairman of ASU Data Enterprise. “This stands as a testomony to ASU’s unwavering dedication to advancing data and fostering groundbreaking analysis. Serving as a cornerstone of infrastructure, Sol propels us in the direction of a future the place new frontiers of discovery and transformative options flourish.”

Since 1993, the Top500 has listed essentially the most highly effective laptop programs biannually by their pace, quantified by the variety of floating-point operations a pc can course of per second, or flops. Sol, which may course of 2.272 petaflops, or 2.272 quadrillion flops, ranks extremely amongst tutorial establishments, clocking in quicker than Mississippi State College’s Hercules, New York College’s Greene and College of Minnesota’s Agate.

“Sol is nearly 2,000 occasions quicker or extra succesful than a contemporary laptop computer, and when it comes to information storage, you could possibly say about 4,000 occasions that,” says Douglas Jennewein, senior director of ASU Analysis Computing.

Whereas customary computer systems use a single central processing unit (CPU) to course of operations, supercomputers comprise many CPUs. This permits supercomputers to attract on the mixed processing energy of many CPUs — clustered collectively as nodes — to sort out large-scale, advanced issues past the scope of a standard machine. Sol possesses 18,000 CPU cores spanning 178 nodes.

To realize this degree of processing and computing energy, ASU drew on its shut relationship with Dell Applied sciences. As a Dell Applied sciences Excessive Efficiency Computing and Synthetic Intelligence Heart of Excellence, ASU labored with the pc and software program big to design and fabricate the brand new supercomputer.

“Endeavoring to search out solutions to the world’s hardest questions requires an unbelievable quantity of computing efficiency. ASU’s Sol supercomputer is designed to just do that, powered by Dell PowerEdge servers and HPC Validated Design Options,” says Dave Lincoln, vice chairman of laptop programs, options and portfolio administration at Dell Applied sciences. “Constructing on the Dell HPC and AI Heart of Excellence with ASU, Sol is the newest in a sequence of highly effective, accelerated programs that may assist pace discovery with AI modeling for a variety of areas — from protein folding to climate predictions.”

Since its inception final yr, Sol and different Analysis Computing assets have supported the analysis initiatives of roughly 20% of principal investigators throughout ten schools, together with the Ira A. Fulton Faculties of Engineering, The Faculty of Liberal Arts and Sciences, the Faculty of Well being Options and the W. P. Carey Faculty of Enterprise.

“There appears to be a lot of methods wherein computing can help in analysis. It has actually expanded in thrilling methods just lately, and we’re simply making an attempt to satisfy that want,” says Gil Speyer, director of the Computational Analysis Accelerator.

Via ASU’s Computational Analysis Accelerator, Speyer works with researchers throughout ASU to establish alternatives the place high-performance computing and software program options can bolster and optimize initiatives.

Supporting an array of disciplines, Sol’s capabilities have been used within the Faculty of Geographical Sciences and City Planning to symbolize large-scale local weather fashions; within the Faculty of Earth and House Exploration to find out the quantity of hydrated materials inside Martian rocks and soils; and within the Faculty of Computing and Augmented Intelligence to develop deep studying fashions to reinforce PET scans for early detection of Alzheimer’s illness.

In accordance with Jennewein, Analysis Computing’s assets have enabled over $170 million in analysis and grant funding over the previous yr. With Sol’s new rating, that quantity is poised to develop.

“Sol is offering a basis for our researchers to perform nice issues at a a lot better scale,” says Sean Dudley, affiliate vice chairman and chief analysis info officer of Data Enterprise. “Specifically, this supercomputer is basically designed to assist each the use and improvement of synthetic intelligence. I admire the chance to offer this spectacular new useful resource to our analysis group and am excited for the work this new system will assist.”

As a result of rising recognition and utility of synthetic intelligence (AI) platforms, Sol was objective constructed to assist a bigger workload of AI applications. This development equips the college with the instruments essential to conduct analysis at a bigger and extra aggressive scale, notably because the college expands its efforts in well being care.

“From imaging information evaluation to genetic analysis, predictive well being fashions and drug discovery, superior computing programs will play an important position in supporting ASU’s new medical college,” Jennewein says.

Sol additionally expands the college’s potential to not solely assist however join researchers throughout disciplines together with engineering, enterprise, well being sciences and extra. By being aggressive on a worldwide scale, it raises the caliber of the analysis itself.

“The impact of investing in an instrument like this on our campus is {that a} group of apply emerges,” Speyer says. “It appears to carry individuals collectively and community them in such a method that they develop into, as a group, extra aggressive research-wise.”

ASU researchers concerned about exploring how Sol can advance their analysis can request entry by way of Analysis Computing.