How Supercomputers Evolved: From Early SGI Systems to China’s Exascale Machines
This article traces the evolution of global supercomputing—from early US and Japanese initiatives to Europe’s coordinated investments—and details China’s rapid development of successive supercomputer generations, highlighting landmark systems such as SGI Power Challenge XL, Dawning‑2000, DeepComp series, the “元” platform and the “东方” machine, as well as homegrown high‑performance software like HPSEPS and HPLES.
Global Supercomputing Landscape
Supercomputing is a strategic pillar of national scientific capability, supporting large‑scale scientific projects, economic development, and defense.
United States
US agencies such as DOE, NSF and NASA have driven multiple strategic programs (SCP, HPCC, ASCI, PACI, HPCS) that keep the US at the forefront of high‑performance computing, exemplified by Frontier’s TOP500 lead in 2022.
Japan
RIKEN’s investment in the “Fugaku” successor “Fugaku NEXT” aims to deliver a system up to ten times the performance of Fugaku by 2030.
Europe
European initiatives DEISA, PRACE and the EuroHPC Joint Undertaking have funded billions of euros to build eight world‑class supercomputers between 2018‑2026.
China’s Supercomputing Evolution
Since the mid‑1990s China has introduced successive generations of supercomputers, beginning with the SGI Power Challenge XL (1996) and Hitachi SR2201 (1998), followed by indigenous systems such as Dawning‑2000 (2000), DeepComp 6800 (2003) and DeepComp 7000 (2008). The “元” platform (2014‑2018) achieved 2.69 PFLOPS with a hybrid CPU‑GPU architecture, and the “东方” system (2019) provides 100 PB storage and 200 Gb HDR InfiniBand connectivity.
Key Domestic Systems
SGI Power Challenge XL – MIPS R10000 CPUs, 64 GFLOPS.
Hitachi SR2201 – PA‑RISC CPUs, 96 GFLOPS.
Dawning‑2000 – 111.7 GFLOPS, first Chinese system surpassing 10 TFLOPS.
DeepComp 6800/7000 – up to 102.8 TFLOPS, TOP500 rank 14 (2008) and 19 (2008).
元 – 2.69 PFLOPS (CPU 750 TFLOPS, GPU 1.94 PFLOPS), 152 TB RAM, 6.9 PB storage.
东方 – 100 PB storage, 200 Gb HDR InfiniBand network.
Software Innovations
The Chinese Academy of Sciences Supercomputing Center has developed high‑performance software stacks, including the HPSEPS package for dense and sparse eigenvalue problems and the HPLES framework for multi‑preconditioned sparse linear algebra, supporting applications in fluid dynamics, multi‑physics simulation and large‑scale multiphase flow.
Signed-in readers can open the original source through BestHub's protected redirect.
This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactand we will review it promptly.
Architects' Tech Alliance
Sharing project experiences, insights into cutting-edge architectures, focusing on cloud computing, microservices, big data, hyper-convergence, storage, data protection, artificial intelligence, industry practices and solutions.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.