More than 63,000 Xeon GPU Max chips and 21,000 Xeon CPU Max chips

Intel & Argonne National Laboratory announces the successful installation of a blade in the Aurora supercomputer, bringing it one step closer to full functionality.

Intel-based Aurora supercomputer boasts 2 ExaFLOPS computing power, potentially surpassing AMD’s boundary

The Aurora supercomputer has been the victim of several delays since its creation, but we could finally see it working. For those unaware, the Aurora supercomputer incorporates Intel’s Xeon CPU Max and Xeon GPU Max series, raising its performance to 2 ExaFLOPS. One of the applications of the Aurora platform will be to provide a state-of-the-art generative AI model for science.

It offers 10,624 nodes with 21,248 Sapphire-Rapid SP line Xeon processors. It comes with a total of 63,744 GPUs based on the Ponte Vecchio design, allowing it to deliver a maximum injection of 2.12 PB/s and a maximum bisection bandwidth of 0.69 PB/s.

Here’s how the Intel-powered Aurora supercomputer has an edge, as Intel Super Compute Group Vice President Jeff McVeigh previously explained:

  • Intel Data Center GPU Max Series Outperforms Nvidia H100 PCIe Card by 30% on Average on Various Workloads1, While Independent Software Vendor Ansys Shows 50% Speedup for Max Series GPU vs. H100 on Applications AI-accelerated HPC.
  • The Xeon Max Series processor, the only x86 processor with high-bandwidth memory, delivers a 65% improvement over AMD’s Genoa processor on the High Performance Conjugate Gradients (HPCG) benchmark1, using less memory ‘energy. High memory bandwidth was noted as one of the most desired features by HPC customers.
  • 4th Gen Intel Xeon Scalable processors – the most widely used in HPC – deliver an average 50% speedup over AMD’s Milan4, and energy company BP’s latest 4th Gen Xeon HPC cluster delivers an 8x increase in performance over its previous generation processors with improved power efficiency.
  • The Gaudi2 Deep Learning Accelerator performs competitively on deep learning training and inference, with performance up to 2.4x faster than Nvidia A100.

For memory capacity, the Aurora supercomputer has 10.9 PB of DDR5 system DRAM, 1.36 PB of HBM capacity through CPUs, and 8.16 PB of HBM capacity through GPUs. Moreover, it uses an arrangement of 1,024 storage nodes offering a total capacity of 220TB. If you are curious how this gigantic system will be used, here is a brief explanation:

From tackling climate change to finding cures for deadly diseases, researchers face monumental challenges that require advanced computing technologies at scale. Aurora is poised to meet the needs of the HPC and AI communities, providing the tools needed to push the boundaries of scientific exploration.

The latest Intel Data Center GPU Max Series 1550 processor, running on Aurora, delivers the best SimpleFOMP performance, beating the NVIDIA A100 and AMD Instinct MI250X accelerators. However, the supercomputer has yet to pass preliminary tests. After that, it should appear in the list, potentially surpassing the AMD-powered Frontier supercomputer. The Aurora supercomputer is on track to be fully functional by this year.

Leave a Comment