Hewlett Packard Enterprise, not to be confused with the personal computing brand it split from in 2015, said today that it acquired Cray, an iconic supercomputer manufacturer with a rich history in the computer industry. It is said that the agreement is valued at $ 1.4 billion.
Cray, which was founded in 1972 by "the father of supercomputing," Seymour Cray, is currently hired to build two of the world's fastest supercomputers for two US Department of Energy laboratories. US: the Oak Ridge National Laboratory and the Argonne National Laboratory. Both systems, one called Frontier that is being built in partnership with AMD and one called Aurora with Intel, have the promise of delivering the so-called "ex-scale" performance, with a raw performance power of more than 1.5 exfaflops, or a quintillion of calculations per second.
Such exascale supercomputer systems do not yet exist, and it makes sense that HPE would like to have its own participation in the next "exascale" opportunities ", for which the company says there is a growing market of mainly government contracts, which is currently estimated at $ 4 billion in the next half decade. Currently, IBM is the manufacturer of the two fastest supercomputers in the world, Summit and Sierra.
"The answers to some of society's most pressing challenges are buried in huge amounts of data," HPE Chief Executive Antonio Neri said in a statement. "Only by processing and analyzing this data can we unlock the answers to the critical challenges of medicine, climate change, space and more. Cray is a world leader in supercomputing technology and shares our deep commitment to innovation. By combining our equipment and world-class technology, we will have the opportunity to drive the next generation of high-performance computing and play an important role in advancing the way people live and work. "
The supercomputers of this scale can be enormously beneficial to data-intensive industries such as astronomy, climate science, medicine, neuroscience and physics, and increasingly, these systems can be used in artificial intelligence research, which, in turn, It can help accelerate many other areas of scientific research.That said, a supercomputer like Aurora or Frontier tends to be built and financed only by the government for, at least initially, military applications.
USA, China, Japan and other countries are currently in a race to be the first to reach an exscaled performance, with China as the favorite to reach and The milestone in 2020. The second fastest unit in the United States, Sierra, is also scheduled to open the air next year (disconnected from the outside world), so it can be used to manage the country's nuclear arsenal.
. That does not mean that supercomputers are not being used for more important scientific research. The Theta from the Argonne National Laboratory, for example, is currently being used to conduct cutting-edge neuroscience research involving the mapping of rodent brains. But just like the science-based races that saw the first satellites and, eventually, the manned space missions, the type of raw reference search with the universal-scale race is usually reserved for international competition among federal agencies.
Fortunately, over time, these next extreme-scale supercomputers will probably be freed from the military apparatus and get to work guessing new ideas from data. According to HPE, the acquisition of Cray is primarily to help it gain an advantage in AI research and the hardware needed to train ever larger neural networks.