Nvidia will build the 'world's fastest AI supercomputer'

Nvidia and Cineca, an Italian inter-university consortium and main supercomputing middle, have introduced plans to build ‘the world’s fastest AI supercomputer.’

The upcoming Leonardo system will use almost 14,000 Nvidia A100 GPUs for a wide range of high-performance computing duties. The peak efficiency of the system is anticipated to hit 10 FP16 ExaFLOPS. 

The supercomputer will be based mostly on Atos’ BullSequana XH2000 supercomputer nodes, every carrying one unknown Intel Xeon processor, 4 Nvidia A100 GPUs and a Mellanox HDR 200Gb/s InfiniBand card for connectivity. The blades are water cooled and there are 32 of them in every HPC cupboard.  

(Image credit score: Atos)

The BullSequana XH2000 structure may be very versatile, so it could possibly home any CPU and GPU and, to that finish, we will solely guess which Intel Xeon processor will be used for Leonardo.

(Image credit score: Atos)

Scientists from Italian universities plan to make use of Leonardo for drug discovery, house exploration and analysis, and climate modelling. 


By admin

Leave a Reply

Your email address will not be published. Required fields are marked *