HPE just announced its new high-performance computing (HPC) and artificial intelligence (AI) infrastructure portfolio that includes world-class HPE Cray supercomputing solutions and two systems optimized for training large language models (LLM), processing natural language (NLP) and multimodal training system. The new supercomputing solutions are designed to help global customers accelerate scientific research and innovation.
As stated by Trish Damkroger, senior vice president and general manager of HPC and AI Infrastructure Solutions at HPE, the reason for this renewal is based on the fact that “service providers and countries that are investing in sovereign AI initiatives are increasingly turning to “Increasingly to high-performance computing as the fundamental backbone that enables large-scale AI training, accelerating discovery and innovation.” For this reason, it is considered that the new solutions will allow companies to speed up the deployment of AI systems and obtain value more quickly and efficiently.
High performance computing solutions
The North American multinational has developed a complete and extensive portfolio of high-performance computing products and solutions. The new developments are based on the HPE Cray Supercomputing EX system and have been designed for research institutions tasked with solving the biggest global challenges. The proposal is based on the industry’s first 100% fanless direct liquid cooling (DLC) architecture and spans all layers of HPE supercomputing solutions, including compute nodes, networking and storage, complemented by a new offering of software. Among all the solutions, the following stand out:
· HPE Cray Supercomputing EX4252 Gen 2 Compute Blade – Capable of delivering up to 98,304 cores in a single enclosure, the HPE Cray Supercomputing EX4252 Gen 2 Compute Blade provides the most powerful single rack unit system available for supercomputing. Featuring eight 5th Generation AMD EPYC processors, this compute blade offers the advantage of high CPU density, allowing customers to get superior performance in the same space. HPE Cray Supercomputing EX4252 Gen 2 Compute Blade will be available in Spring 2025.
· HPE Cray Supercomputing EX154n Accelerator Blade – To dramatically reduce the time required to complete a supercomputing workload, the HPE Cray Supercomputing EX154n Accelerator Blade can house up to 224 NVIDIA Blackwell GPUs in a single enclosure. Equipped with the NVIDIA GB200 Grace Blackwell NVL4 Superchip, each acceleration blade includes four Blackwell GPUs connected via NVIDIA NVLink and unified with two NVIDIA Grace CPUs and four NVIDIA B200 NVLink-C2C GPUs. The HPE Cray Supercomputing EX154n Accelerator Blade is expected to be generally available in late 2025.
· HPE Slingshot interconnect 400 – The next generation of HPE’s exascale-capable interconnect portfolio delivers network interface controllers (NICs), cables and switches at speeds of 400 gigabits per second. HPE Slingshot interconnect 400 doubles the line speed over the previous generation, offering features such as automatic congestion management and adaptive routing for ultra-low tail latency, allowing customers to run large workloads with significantly less infrastructure. grid. This version of HPE Slingshot will be available for clusters based on HPE Cray Supercomputing EX systems starting in Fall 2025.
· HPE Cray Supercomputing Storage Systems E2000 – This high-performance storage system, designed for large-scale supercomputers, operates delivering more than double levels of input/output (I/O) performance compared to the previous generation. HPE Cray Supercomputing Storage Systems E2000 is based on the open source Luster file system and enables better utilization of both CPU and GPU compute nodes by reducing idle time during I/O operations. This HPC storage system will be generally available on HPE Cray Supercomputing EX systems in early 2025.
· HPE Cray Supercomputing User Services Software – HPE introduces a new software offering that optimizes computing in the execution of intensive workloads to improve the user experience. Available now, HPE Cray Supercomputing User Services Software includes features that help customers optimize system efficiency, regulate power consumption, and flexibly run diverse workloads across supercomputing infrastructure.
