Hewlett Packard Enterprise has announced a new collaborative effort with the Stephen Hawking Center for Theoretical Cosmology. The new effort, which leverages HP Superdome Flex servers, is a step forward for the COSMOS project, which has existed in one form or another since 1997. Back then, SGI and Intel were the main players, although HPE seems to be addressing the updates and improvements by itself so far.
"Our COSMOS group is working to understand how space and time work, from before the first trillion trillion second after the Big Bang to today," said Hawking, the Tsui Wong. Avery Director of Research in the Department of Applied Mathematics and Theoretical Physics of Cambridge, in a statement. "The recent discovery of gravitational waves offers incredible insights into black holes and the entire Universe With new and exciting data like this, we need flexible and powerful computer systems to move forward and test our theories and explore new concepts in fundamental physics" .
HPE says that this new Superdome will join an existing HPE Apollo supercomputer, but I will hardly handle heavy work by itself. COSMOS already has an SGI Altix UV2000, Cosmos2 (1536 Xeon E5-4650L) and Cosmic (288x Intel Xeon E5-4650L, along with 24 Intel Xeon Phi MIC (5110P).
The staff of the Center for Theoretical Cosmology
HPE Superdome Flex is undoubtedly a beast: the machine can scale from 4 to 32 takes and supports 768 GB to 48 TB of memory Exactly what benefits it will bring to the project Cosmos – Although HPE does much of its computing on memory (which means that it contains large data sets completely in DRAM), it is not clear how much of a specific benefit it provides NextPlatform has made a deep immersion in memory computing, targeting Although it is unquestionably good for solving certain types of problems, it is not the panacea that HP says: "The simplistic way of saying it is that memory works if everything fits, and when it does not fit, it really does not fit and you have a problem", NextPla tform w rites.
Superdome Flex. Credit: HPE
DRAM also costs much more per byte than conventional storage and consumes much more energy than NAND flash memory or magnetic disks. It is also not clear whether the COSMOS project can make use of the platform's in-memory computing if other systems on the same network are not optimized to do so as well. On the other hand, one of the main differences between an HPC cluster and a conventional platform is that it is often worth carrying out extensive customization in a system for HPC workloads, thanks to the long-term improvements that such optimization can provide.
The new Superdome Flex system will not be limited to COSMOS, but will also be available to many other research departments.
"High performance computing has become the third pillar of research and we look forward to new developments in the mathematical sciences in such diverse areas as oceans modeling, medical imaging and soft matter physics," said the professor. Nigel Peake, Head of the Cambridge Department of Applied Mathematics and Theoretical Physics.
Top image: NASA / JPL-Caltech