Supercomputers in 2022: key developments

Developments in supercomputers began back in 1964 with the CDC 6600 and, since then, there have been great advances in hardware and software that have led to these machines having a hundred times more power. The research has led scientists to build quantum computing platforms, with India included in the top 100 supercomputers list for many years.

In 2022, there were plenty of supercomputer announcements, with some new chips powering computers, new hardware now attached to machines to boost performance, and all-new supercomputers.

Here is the list of major developments in supercomputers that happened in 2022.

PARAM Supercomputers

In January, the Indian Institute of Science, Karnataka, launched India’s most powerful supercomputer, param pravego designed by the Center for the Development of Advanced Computing. The computer has 4 petabytes of storage and 3.3 petaflops of peak computing power. The supercomputer uses Intel Xeon Cascade Lake processors and NVIDIA Tesla V100 cards. PARAM (Parallel Machine) has been developing various supercomputers since its inception in 1987 such as PARAM 8000 which was the first Indian supercomputer, Siddhi-AI, Shivay, Sanganak among many others. India has also launched Param Ganga, its first petascale supercomputer at IIT Roorkee.

Read: A complete list of Indian PARAM supercomputers

tesla dojo

On Tesla AI Day, Elon Musk made many announcements, from self-driving cars to Optimus Robot, but the news about Dojo and ExaPod were the highlights. The company is exploring the opportunity to sell computing time at its Dojo like Microsoft on Azure and Amazon through AWS. Tesla demonstrated that Dojo has been outperforming other GPUs in occupancy networking and auto-labeling. Previously, Tesla was using 5760 NVIDIA A100 GPUs to build FSD models, which ranked fifth in the world with 1.8 exaflops of power.

andromeda brains

American AI startup, Cerebras System, unveiled one of the largest AI supercomputers called Andromeda. The supercomputer is basically made for academic and commercial research. The supercomputer is capable of performing a quintillion operations per second and also 1 exaflop of AI computing in a 16-bit floating point format. One of the highlights of its launch was that the supercomputer was built for less than $35 million, almost 10 times less than Frontier, one of the largest supercomputers at Oak Ridge National Laboratory.

NVIDIA and Microsoft supercomputer

In November, NVIDIA announced that they are partner with Microsoft to build the most powerful AI supercomputer using Microsoft’s Azure platform to build the infrastructure. The die will be comprised of NVIDIA’s H100 and A100 GPUs and will also include their Quantum-2 400Gb/s Infiniband for super-network performance. Also, in its latest announcement, NVIDIA’s H100 GPU has pumped power on its High Performance Computing (HPC) platform by enabling a complete technology stack for AI and HPC software.

Read: The 7 Best AI Hardware Released in 2022

Frontier with AMD technology

Frontier, the Oak Ridge National Laboratory’s supercomputer powered by AMD, ranked first on the Top500 list of the world’s fastest supercomputers. It is also the first true exascale system with an HPL rating of 1.102 exaflop/s. The place was taken over by the Fugaku system at the RIKEN Center for Computer Science in Japan.

Meta AI Supercomputer

In January, Meta launched AI Research Supercluster (RSC), claiming that it is one of the fastest AI supercomputers in the world. It is expected to work with hundreds of languages, analyze text and images to build better AI models. In addition to computer vision and extensive language tasks, Meta said that CSR will also be essential in building the foundation for the metaverse. To build the supercomputer, Meta partnered with NVIDIA to use 760 of their DGX A100 systems that come with 6080 A100 GPUs linked on Quantum 200Gb/s InfiniBand to deliver 1895 petaflops of TF32 performance.

intel aurora

In partnership with the US Department of Energy and Argonne National Laboratory, Intel is developing the Exascale Aurora Supercomputer. Intel will use its Max Series family of products to provide HPC and AI computing. The supercomputer is expected to house 10,000 blades, containing six Max-series GPUs and two Xeon Max CPUs when it comes online in 2023. In addition, Intel is also delivering Xeon Max CPUs to Los Alamos National Laboratory and Kyoto University. for the development of supercomputers. Xeon Max is the first and only x86-based processor with 56 performance cores and uses 68% less power than AMD Milan-X.

Leave a Reply

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