Context
Japan’s Fujitsu Laboratories has developed an AI model to predict tsunami flooding in coastal areas in real-time. The technology company used Fugaku, the world’s fastest supercomputer, to develop the model.
About the AI model
- The team of researchers generated training data for 20,000 possible tsunami scenarios based on high-resolution tsunami simulations, using the supercomputer.
- They created the AI model using these data sets.
- In the event of an earthquake by using the data of tsunami observed offshore helps predict flooding in coastal areas before the wave makes landfall.
Significance of the model
- This will make it possible to accurately and rapidly obtain flooding forecasts for specific areas.
- The system is also said to make evacuation measures more efficient.
- The model can also be run in seconds on ordinary PCs, making it easier to build practical, real-time systems, which previously required supercomputers.
About Fugaku supercomputer
- The Fugaku supercomputer was developed in Japan in 2014 at the RIKEN Center for Computational Science.
- The supercomputer was termed the world’s fastest in June 2020 in the TOP500project which ranks and details the 500 most powerful non-distributed computer systems in the world according to their LINPACK benchmark results.
- It has a speed of 442 Peta Flops.
|
What are Supercomputers?
- Supercomputers are a class of extremely powerful computers.
- The term is commonly applied to the fastest high-performance systems available at any given time.
- The performance of a supercomputer is commonly measured in floating-point operations per second (FLOPS).
- Most modern supercomputers use a Linux-based operating system.
- Unlike conventional computers, they usually have more than one CPU (central processing unit).
- They can operate on pairs of lists of numbers rather than on mere pairs of numbers.
- Most supercomputers have a very large storage capacity, as well as a very fast input/output capability.
Indian Supercomputers
- Two supercomputers from India — PARAM Siddhi-AI and Mihir.
- They have made it to the TOP500 list of the world’s most powerful supercomputers.
- The two systems have been ranked 63rd and 146th, respectively.
- PARAM Siddhi-AI: PARAM Siddhi-AI has a peak capacity of 5.267 Petaflops and an Rmax capacity of 4.6 Petaflops. It was led and built on NVIDIA DGX SuperPOD architecture.
- Mihir’ is a 2.8 Petaflop supercomputer.
|
