Recently, NVIDIA officially unveiled the next generation GPU architecture “Ampere”, one of the first products is AI artificial intelligence and ML machine learning system DGX A100, equipped with up to eight ampere architecture A100 GPU, and a very low-key two AMD 64 core Dragon processor. Today, AMD officially unveiled the answer, two top-of-the-line Dragon 7742, the same 64 core 128 thread, the base frequency is slightly lower than 2.25GHz, but the maximum acceleration is 3.4GHz, while the thermal design power consumption control at 225W – the official price of $6950.
AMD’s notification emphasizes that the second generation of Dragon processors is the world’s first and only x86 architecture server processor that supports the PCIe 4.0 bus, providing up to 128 PCIe 4.0 channels per processor to support high-performance computing and high-speed connectivity to other devices such as GPUs.
According to the data, the NVIDIA DGX A100 system is also equipped with PCIe 4.0 NVMe SSD with up to 15TB capacity, but the specific brand and product information is still unclear.
With superior specifications and performance, relatively low prices, AMD II Is a lot of high-performance computing systems, especially the Dragon 7742 can be a star product, especially the world’s super-fast two billion sub-prime supercomputers, EI Frontier Captian, have the second generation of Dragon, and AMD Radeon Instinct acceleration card.
Other supercomputers for AMD II Dragon are:
– HLRS “HAWK”:
The Stuttgart High Performance Computing Center (HLRS) has selected the DDN EXAScaler and its IME scale-out flash memory system to support the new flagship supercomputer system Hawk.
Hawk enables scientists and engineers to study larger, more complex phenomena while supporting data-intensive workflows that combine big data analytics, deep learning, and simulation.
Hawk is part of the HPE Apollo 9000 system, one of the world’s ultra-fast supercomputers, and one of the most versatile systems in Europe for scientific and industrial computing.
– BIG RED 200:
Indiana University has deployed the Big Red 200, one of the nation’s top supercomputers, to support advanced research in artificial intelligence, machine learning, data analytics, science and medical research. The system replaces the highly successful Big Red 200 II Cray supercomputer, which operates more than 60 000 billion times per second. In addition to AMD II Dragon, it is also equipped with the Tensor Core GPU with NVIDIA Ampere architecture.
– NEC SX-AURORA TSUBASA:
German weather service Deutsche Wetterdienst (DWD) has chosen NEC to build the 50 million euro vector supercomputer, which is expected to be six times faster than DWD’s current supercomputer and more powerful than the supercomputers used by the Met Office in the UK and France.
NEC SX-Aurora TSUBASA supports monitoring, analysis and prediction in the process of meteorological and climate change, as does complex forecasting and early warning services in the event of severe climate events.
– ETH ZURICH EULER VI:
The Federal Institute of Technology in Zurich (ETH Zurich) uses the second generation AMD EPYC 7742 processor in its Euler VI system.
– Dell EXPANSE:
The San Diego Supercomputer Center (SDSC) has chosen Dell to build its next-generation supercomputer, Expanse, and is expected to deploy in mid-2020, promising to deliver the highest performance of 5 petaflops (trillionfloat slotted-per-second operations).
The system will help thousands of researchers accelerate the process of discovery in the fields of astronomy, molecular dynamics, machine learning and artificial intelligence.
– CRAY ARCHER2:
Cray and HPE have been awarded 48 million euros to expand the high-performance computing power of the UK’s National Research and Innovation Agency (UKRI) with Cray’s next-generation Shasta supercomputer.
THE ARCHER2 SUPERCOMPUTER WILL BE EQUIPPED WITH SLINGSHOT INTERCONNECT, CRAY CLUSTERSTOR HIGH-PERFORMANCE STORAGE, THE CRAY SHASTA SOFTWARE PLATFORM, AND THE SECOND GENERATION AMD EPYC PROCESSOR, WHICH WILL BE 11 TIMES THE PERFORMANCE OF ITS PREDECESSOR, ARCHER.
ARCHER2 will be the first Shasta system announced at EMEA (European Middle East and Africa) and the second in the world for academic research.
– CRAY VULCAN:
The British Atomic Weapons Institute (AWE) chose Cray, HPE, and used the Cray Shasta supercomputer to keep Britain safe.
The AWE Vulcan supercomputer will be equipped with a Shasta supercomputer with performance expected to exceed 7PFlops, with architectures including Cray Slingshot Interconnect, second-generation AMD EPYC processor, Cray ClusterStor Luster storage, and I/O performance of nearly 100GB per second.
France’s National High Performance Computing Organization (GENCI) and Atos announced the world’s first operational supercomputer, Joliot-Curie, based on the Atos BullSequana XH2000 solution, with continuous performance improvements that have so far exceeded 5PFlops, ranking 59th in the world.
– ATOS BULLSEQUANA XH2000:
The French Meteorological Agency (M?t?o-France) has selected two BullSequana XH2000 supercomputers from Atos, based on a second-generation AMD EPYC processor, which can more than five times more computing power than current solutions and help it make more scientific breakthroughs in weather forecasting.
Each new supercomputer can handle more than 1 trillion operations per second and is one of the world’s most powerful supercomputers for meteorological research.
– CRAY HPC11:
The AFLCMC’s Joint Oak Ridge National Laboratory (ORNL) procured cray HPC11, a Cray Shasta supercomputing system provided by Cray, under a contract value of $25 million.
The HPC11 supercomputer will provide more accurate weather forecasts for the actions of the U.S. Air Force and Army around the world. The HPC11 will be one of the first supercomputing systems for military weather forecasting with a second-generation AMD EPYC processor.