According to the list, the number of supercomputers deployed in China continues to rank first in the world. Among the top 500 supercomputers, 226 are deployed by Chinese customers, accounting for more than 45% of the total. Lenovo, dawn and Inspur, the Chinese manufacturers, are the top three supercomputer suppliers in the world, delivering 312 supercomputers, accounting for more than 62% of the total. Lenovo has contributed more than 3.55 billion times of supercomputing, ranking second in the world.
In terms of single floating-point operation, the supercomputing supermuc-ng designed and manufactured by Lenovo for Leibniz Supercomputing Center in Germany ranks 13th, with a certified computing capacity of 190 million floating-point operations per second (19476 PFlops). Lenovo also has the worlds leading level in the research and development of ultra-high performance supercomputing.
In the list of top 100 supercomputers released at the end of last year, Lenovo has 39 of the top 100 supercomputers in China, ranking first among manufacturers. Among the top ten supercomputers in China, except for the cluster built by the national concurrent computer engineering technology research center and the National University of Defense Science and technology, the rest are all developed and deployed by Lenovo.
Fugaku is the worlds first arm architecture processor driven top 500 high-performance computing cluster, with a certified computing power of more than 5.13 billion times per second.
Last year, Lenovo has joined forces with some well-known enterprises in the field of HPC to establish the e-level computing Vision Committee, which is dedicated to solving technical problems and promoting the e-level, or 10 billion times per second computing power supercomputers to the world.
E-level computing is the HPC technology peak that big countries including China, the United States and Japan are pursuing, which involves all aspects of HPC system design, including infrastructure, alternative cooling technology, efficiency, density, rack, storage, integration of traditional HPC and AI, etc. (quiet)