HPE Arm-based systems with leading Fujitsu A64FX processors bring GPU performance to CPUs. This breakthrough architecture delivers HBM performance to simulation applications and readies you for the future.
We see many organizations seeking step-function performance improvements and considering GPUs as accelerators to power their high performance computing (HPC) environments. Many struggle or decline to transform. The reason is that a majority of in-house codes were developed for CPU-based compute.
Unless new solutions are compatible with existing simulation applications, organizations will have to refactor large volumes of source code to run on the specialized hardware. The cost and complexity of rewriting code for accelerators takes considerable effort, and the entire process can be extremely iterative and time-consuming. This slows down time to value and leaves critical workers such as scientists, engineers, and researchers racing to make up time, money, and resources.
Instead, more organizations are putting their software engineering effort into advancing the capabilities of their existing simulation applications, to convert their intellectual property into competitive advantage. The good news is, they can unleash the performance of GPUs with CPU-based computation and bypass the difficulties of complex refactoring projects.
Breakthrough systems are providing exceptional compute power and scalability, leveraging CPUs that are purpose-built for HPC. This new breed of processor delivers world-leading performance and high bandwidth memory (HBM) bandwidth to run complex and data-intensive simulations. HBM overcomes the memory bandwidth limitation that hinders application performance. This helps organizations to modernize infrastructure and boost their return on investment for existing applications without having to refactor or rewrite their codes for separate accelerators.
Industries are using the latest advancements in computing to take on increasingly bigger scientific, technological, and business-related problems. From diverse areas such as manufacturing, seismic interpretation, healthcare and life sciences, and climate research, HPC is key to optimizing the most demanding simulation applications, accelerating innovation, and delivering insights when and where they are needed most. This is how companies stay one step ahead of the competition.
To develop and maintain leadership capabilities requires experience with diverse computing architectures while simultaneously developing core HPC application capability. The most successful organizations have a strategic plan for HPC adoption that helps them avoid common roadblocks, so they can enhance their technology infrastructure quickly and keep moving forward.
Game-changing CPU performance to accelerate HPC innovation
Engineering, science, and research organizations are achieving greater productivity and faster insight with an architecture that is designed specifically for HPC. HPE continues to deliver the world’s fastest Arm-based systems powered by leading Fujitsu A64FX™processors. Together, our robust solutions are helping organizations achieve more value with in-house codes, bringing new HPC capabilities to floating point-intensive workflows where performance is impacted by memory bandwidth.
HPE Apollo 80 Arm-based systems bring the benefits of GPU computational performance to the CPU without the lifting costs of making major changes to application codes. Porting codes to the HPE Apollo 80 is relatively simple and inexpensive. With a CPU that is accessible to their current applications, organizations benefit from an agile technology infrastructure with an insatiable appetite for data, enabling them to run sophisticated simulations for more immediate and accurate results. By rebalancing the architecture using the A64FX with direct-attached HBM, the CPU receives data at a breakneck rate instead of waiting for data to compute.
HPE Apollo 80 systems also bring Scalable Vector Extensions (SVE) to HPC for the first time, which are engineered for HPC software codes and can vectorize elaborate structures—this includes gather-load, scatter-store, and horizontal reductions—all while allowing for efficient and flexible vectorization of codes. These groundbreaking technologies provide a rapid path for deploying and tuning simulation applications, so users can focus on improving their simulations and driving better business outcomes rather than on rewriting code.
By rebalancing the architecture using this powerful CPU with directly attached HBM, the powerful A64FX CPU receives data at extraordinary rates and is not left waiting for data to compute. This memory-CPU bandwidth matched with Arm SVE enables performance increases without requiring accelerators. The user can spend precious time and talent improving capabilities, and tuning for performance rather than re-writing their code.
To accelerate the rate of innovation, HPE Apollo 80 Arm-based systems come with a comprehensive software stack. HPE Cray Programming Environment is an integrated software suite for software development of HPC applications, and system administration and workload management tools to run jobs efficiently while managing heterogeneous clusters. The programming environment suite includes compilers, tools, and libraries designed to maximize programmer productivity, application scalability, and performance. These capabilities will drive major improvements in science, engineering, and research with the ability to develop unique HPC code and port to the new SVE architecture.
The use of HPC among oil and gas producers in the energy industry is a powerful example. HPC application performance is critical for upstream applications like seismic interpretation and reservoir modelling, mid and downstream applications like fluid dynamics, and emerging predictive applications that are improving operations, analyzing risk, and developing new core competencies critical for “next era” success.
HPE and Fujitsu are revolutionizing the possibilities of simulation applications and preparing for the future of HPC. Don’t miss your opportunity to transform. Contact us today to learn more.
A member of HPE’s High Performance Computing Product Marketing team, Roger has led marketing programs for HPC/AI products, HPE HPC Software, and Arm in HPC. Roger’s background includes HPC hardware and software marketing, industry marketing for engineering and science, alliance management and product marketing from supercomputing to departmental HPC. In his spare time Roger is an enthusiastic coach who loves sharing his lifelong passions, welcoming and developing cyclists and skiers of all ages.
Advantage EX Experts
Hewlett Packard Enterprise
twitter.com/hpe_hpc
linkedin.com/showcase/hpe-ai/
hpe.com/info/hpc
AdvEXperts
Our team of Hewlett Packard Enterprise Advantage EX experts helps you dive deep into high performance computing and supercomputing topics.