Intel has recently unveiled the full specifications of the Aurora supercomputer, which is being touted as one of the most advanced supercomputers ever built. The Aurora supercomputer boasts an impressive array of features, including thousands of CPUs and GPUs, top-of-the-line memory capacity, and a vast array of deep learning accelerators. In this article, we will delve into the specifications of the Aurora supercomputer and explore its various features.
Aurora Supercomputer
The Aurora supercomputer is equipped with 21,248 Xeon CPUs and 63,744 GPUs based on the Ponte Vecchio design. This powerful combination offers a peak injection of 2.12 PB/s and a peak bisection bandwidth of 0.69 PB/s. The CPU and GPU combination in the Aurora supercomputer has been designed to provide users with the ultimate performance and computing power they need to achieve the best results.
Intel Data Center GPU – Max Series
The Aurora supercomputer is also home to the Intel Data Center GPU Max series which outperforms the Nvidia H100 PCIe card by an average of 30% on diverse workloads. This GPU is designed to provide high-performance computing power in diverse fields, including AI, HPC, and data analysis.
Xeon Max Series CPU
The Xeon Max Series CPU is another impressive feature of the Aurora supercomputer. This CPU exhibits a 65% improvement over AMD’s Genoa processor on the High Performance Conjugate Gradients (HPCG) benchmark while using less power. This means that the Aurora supercomputer can handle massive workloads with ease, reducing overall power consumption while maintaining top-level performance.
4th generation Intel Xeon Scalable processors
The Aurora supercomputer is also equipped with 4th Gen Intel Xeon Scalable processors, which offer a 50% average speedup over AMD’s Milan. These processors are designed to deliver exceptional performance and speed, making them ideal for complex workloads that require intelligence, power, and efficiency.
Gaudi2 Deep Learning Accelerator
The Aurora supercomputer is home to the Gaudi2 Deep Learning Accelerator, which performs competitively on deep learning training and inference. This accelerator is up to 2.4x faster than the Nvidia A100 and is designed to provide users with the ultimate deep learning experience. With the Gaudi2 Deep Learning Accelerator, scientists and researchers can explore the boundaries of scientific research without compromising on performance.
The Aurora supercomputer features an impressive 10.9 PB of DDR5 system DRAM, 1.36 PB of HBM capacity through the CPUs, and 8.16 PB of HBM capacity through the GPUs. This vast memory capacity ensures that the supercomputer can store and analyze large volumes of data effectively.
Purpose of Aurora Supercomputer
The Aurora supercomputer is poised to address the needs of the HPC and AI communities, providing the necessary tools to push the boundaries of scientific exploration, enabling researchers to delve deeper into complex problems that require advanced computational power. Designed to provide unparalleled performance and computing power, the Aurora supercomputer serves as an ideal tool for scientists, researchers, and data analysts.
New Intel Data Center GPU Max Series 1550
The newest addition to the Intel Data Center GPU Max series is the 1550 GPU, which provides the best SimpleFOMP performance, beating out the Nvidia A100 and AMD Instinct MI250X accelerators. This GPU is designed for large-scale HPC and AI workloads, ensuring that users can carry out complex computing tasks without compromising on performance.
In conclusion, the Aurora supercomputer is set to revolutionize the field of scientific computing and AI. With its powerful combination of CPUs, GPUs, and deep learning accelerators, it is well-positioned to provide users with the ultimate performance and computing power needed to delve deeper into complex problems that require advanced computational power. With its vast memory capacity, the Aurora supercomputer can store and analyze large volumes of data effectively. The Aurora supercomputer is on track to be fully functional this year, and we can only imagine the wonders that it will bring to the world of scientific exploration.