Juelich Supercomputing Center and IBM take home Hans Meuer Award

Share this post:

At this year’s ISC International Supercomputing Conference, scientists of the Juelich Supercomputing Center and IBM will receive the first Hans Meuer Award for their work Active Memory Cubes (AMC), which uses 3D memory technology to reduce energy consumption. Their application enables computation in the memory module – a promising storage concept for future computing systems such as exascale. 

The experts explored how AMCs, an innovation developed at IBM Research, influence performance and energy efficiency based on scientific simulations programs in the areas of fluid mechanics and elementary particle physics.

AMCs allow for data manipulation in the main memory of a system, which holds the advantage that data does not have to be moved between the CPU and the memory. This data-centric approach to computing improves the overall systems performance and saves energy.

“Today, the movement of data consumes more energy than the calculation itself. We were interested in the efficiency of memory vector processes as they are being used in Active Memory cubes for scientific calculations,” said Professor Dirk Pleiter of the Juelich Supercomputing Center (JSC).

New storage concepts like AMC hold promise for future architectures of supercomputing generations, in particular Exascale systems, a class of supercomputers which exceeds the compute power of today’s systems by a factor of 100 to 1,000.

“Our work with AMCs is an excellent example of IBM’s data-centric approach, one that puts computing power everywhere data resides,” said Dr. Thilo Maurer of the Exascale Innovation Center at the German IBM Development Lab.

“By bringing the central processor and memory technologies together, we were able to perform computation within the memory module — the ‘home’ of the data. This highly efficient approach, already present in several of IBM’s current solutions, minimizes data movement latency and significantly reduces energy consumption.”  

As AMC is a technology concept, the scientists researched the performance of the storage cubes with simulations. Based on fluid simulations according to the Lattice-Boltzmann procedure and on quantum chromo dynamics the experts applied widely used analytics methods which are very likely to run on many high performance systems in the future.

To simulate the necessary calculation steps, the scientists – Thorsten Hater, Paul F. Baumeister, Andrea Nobile and Dirk Pleiter from JSC; Hans Böttiger und Thilo Maurer from IBM Deutschland Research & Development, as well as Jose R. Brunheroto from the IBM Thomas J. Watson Research Center – convert the code into VLIW (Very Long Instruction Word) format of the AMCs and interpreted them on a cycle model. The predictions regarding energy efficiency (floating point operations per watt) outperforms the results of alternative platforms by far.  

The Hans Meuer Award honors the most outstanding research paper submitted to the ISC International Supercomputing Conference’s research paper committee. In its inaugural year, the award is named in memory of the late Dr. Hans Meuer, general chair of the ISC conference from 1986 through 2014, and co-founder of the TOP500 project.

More stories

A new supercomputing-powered weather model may ready us for Exascale

In the U.S. alone, extreme weather caused some 297 deaths and $53.5 billion in economic damage in 2016. Globally, natural disasters caused $175 billion in damage. It’s essential for governments, business and people to receive advance warning of wild weather in order to minimize its impact, yet today the information we get is limited. Current […]

Continue reading

DREAM Challenge results: Can machine learning help improve accuracy in breast cancer screening?

        Breast Cancer is the most common cancer in women. It is estimated that one out of eight women will be diagnosed with breast cancer in their lifetime. The good news is that 99 percent of women whose breast cancer was detected early (stage 1 or 0) survive beyond five years after […]

Continue reading

Computational Neuroscience

New Issue of the IBM Journal of Research and Development   Understanding the brain’s dynamics is of central importance to neuroscience. Our ability to observe, model, and infer from neuroscientific data the principles and mechanisms of brain dynamics determines our ability to understand the brain’s unusual cognitive and behavioral capabilities. Our guest editors, James Kozloski, […]

Continue reading