banner



Barcelona Plans The World's Most Diverse Supercomputer

These days, there are a number of unlike approaches to high-operation computing, systems usually referred to equally supercomputers. Virtually of these systems use a massive number of Xeon processors, just we are starting to see the about interesting new machines run accelerators, such as Nvidia's Tesla or Intel Xeon Phi. There's even some talk that massive ARM-based systems could be effective in the future. But what if y'all could try all of these architectures in 1 location?

That'due south the claiming and promise of the new MareNostrum 4 calculator, which is beingness readied for installation at the Barcelona Supercomputing Centre. The new pattern includes a master system for general-purpose employ based on traditional Xeons, plus three new emerging technology clusters, based on IBM Power and Nvidia, Xeon Phi, and ARM-based calculating. While I was in Barcelona for Mobile World Congress, I had a chance to talk to Sergi Girona, Operations Director for the supercomputing centre, who explained the reasoning behind the four dissimilar clusters.

Girona said the centre's main mission is to provide supercomputing services for Spanish and other European researchers, in improver to industry. As part of this mission the center wants to have at least iii "emerging tech clusters," and so it could test different alternatives.

For the general calculating cluster, Girona says the center chose to a traditional Xeon blueprint because it was easier to migrate applications that run on the current MareNostrum 3, slated to exist disconnected adjacent week. The blueprint also had to fit the existing space, within a chapel. (I visited the centre concluding yr and the current supercomputer a twelvemonth agone.)

The new pattern, to be built past Lenovo, will be based on the new Xeon v5 (Skylake), with 3,456 nodes, each with two sockets, and each chip volition incorporate 24 cores each, for a total theoretical peak performance of xi.14 petaflops per 2nd. Well-nigh cores will have 2GB of memory, but half-dozen pct will have 8GB, for a total of 331.7TB of RAM. Each node will have a 240GB SSD, though somewhen some will have 3D XPoint retentiveness, when that is available. The nodes are to be continued via Intel'south Omni-Path interconnect and 10GB Ethernet. The organisation volition too take six racks of storage from IBM, with 15 petabytes of storage, including a mix of wink and hard disk drives. Overall, the blueprint will take up 62 racks—48 for computing, half-dozen for storage, 6 for networking, and 2 for management. Information technology will fill 120 foursquare meters (making for a very dense environment) and draw ane.3 megawatts of power, up from the one megawatt drawn by the previous design. Operation is expected to begin on July ane.

MareNostrum 1-2-3-4

One thing I institute interesting here is how clearly the move to the new generation demonstrates the progression of technology. The previous generation had a peak operation of about 1 petaflop, and this system should be more than 10 times faster, while using only 30 percent more ability. For comparing, the original MareNostrum supercomputer, installed in 2004, had a tiptop operation of 42 teraflops and used 640 kilowatts of power. (The details of performance improvements over iv generations of MareNostrum are in the nautical chart in a higher place). Girona says this ways that what would have taken a year to run on the MareNostrum 1 can be done in a single day on the new organisation. Pretty impressive.

For emerging technology, the site will have iii new clusters. One volition consist of IBM Power 9 processors and Nvidia GPUs, designed to accept a peak processing capability of over 1.5 Petaflop/s. This cluster will exist built past IBM, and involves the same blazon of design being deployed in the Meridian and Sierra supercomputers, http://forwardthinking.pcmag.com/none/329537-u-due south-getting-two-massive-new-supercomputers which the US Department of Energy has commissioned for the Oak Ridge and Lawrence Livermore National Laboratories equally function of its CORAL Collaboration at Oak Ridge, Argonne, and Lawrence Livermore national labs.

The 2d cluster will be made upwardly of Intel Xeon Phi processors, with Lenovo building a system that uses the forthcoming Knights Colina (KNH) version and OmniPath, with a superlative processing capability over 0.5 Petaflop/southward. This also mimics the American CORAL program, and uses the same processors that will be inside the Aurora supercomputer, deputed by the US Department of Energy for the Argonne National Laboratory.

Finally, a third cluster will be formed of 64-bit ARMv8 processors that Fujitsu will provide in a prototype auto, which is designed to use the aforementioned processors that Fujitsu is developing for a new Japanese system to supplant the current K supercomputer. This too should offer more than 0.5 Petaflop/s of meridian performance. The exact timing for the beginning of operations on the emerging clusters has however to be disclosed, Girona said.

Overall, the system will cost $34 million, in a contract won by IBM and funded by the Castilian government. I major reason for having all iv types of computing on site is research, Girona said. The eye, which employs 450 people in total, has 160 researchers with a focus on reckoner science, including compages and tools. In particular, every bit a member of PRACE (Partnership for Advanced Computing in Europe), BSC is trying to focus on leading performance optimization and parallel computing.

Girona said that BSC wants to influence the development of new technologies, and is planning on using the new machine to clarify what will happen in the future, in particular to make sure that software is ready for whatever compages the adjacent machine—likely to get in in almost iii years—volition take. BSC has long worked on tools for emerging architectures, he noted.

Another topic researchers are considering is whether or non it would be worth developing a European processor for It, likely based on the ARM compages.

Barcelona won't have the fastest supercomputer in the world; that record is currently held by the Chinese, with the Americans and Japanese trying to catch up. Merely MareNostrum 4 will be the most various, and potentially the most interesting.

Source: https://sea.pcmag.com/feature/14449/barcelona-plans-the-worlds-most-diverse-supercomputer

Posted by: ramirezablessairse.blogspot.com

0 Response to "Barcelona Plans The World's Most Diverse Supercomputer"

Post a Comment

Iklan Atas Artikel

Iklan Tengah Artikel 1

Iklan Tengah Artikel 2

Iklan Bawah Artikel