High-performance computing (HPC) has grown to a point where it is a critical component of new technology advancements in academia and a wide array of industries in both the public and private sectors. Scientific research, public health, climate modeling, as well as oil and gas exploration are just a few examples where HPC is the driving force behind new innovations and knowledge discovery.

 

Utilizing the x86-architecture, the AMD EPYC™ processor, brings together high core counts, large memory capacity, extreme memory bandwidth and massive I/O with the right ratios to enable exceptional HPC workload performance.

 

AMD is committed to creating a broad partner ecosystem with collaborative engineering to provide tested and validated solutions that are tuned for specific workloads. As a result, AMD EPYC processors are now certified with software vendors providing some of the most popular HPC solutions. Examples include: computational fluid dynamics (CFD), crash simulation, and finite element analysis (FEA).

 

For computational fluid dynamics (CFD), AMD partnered with ANSYS® to take advantage of the AMD EPYC processor’s ample memory bandwidth to enable exceptional performance with their Fluent® software. ANSYS Fluent is used by the automotive, aerospace, consumer goods, energy, and healthcare industries for modeling flow, turbulence, heat transfer, and reactions in applications ranging from air flow over an aircraft wing to combustion in a furnace.

 

Altair Radioss is a leading structural analysis solver for non-linear problems under dynamic loadings, like automotive crash analysis, drop and impact analysis, terminal ballistics, blast and explosion effects, and high velocity impacts. AMD collaborated with Altair to create an optimized solution for Altair’s PBS Professional, a fast, powerful workload manager designed for HPC clusters, clouds and supercomputers. PBS Professional maximizes the utilization of an AMD EPYC processor cluster and increases the job throughput of Radioss.

 

OpenFOAM®, is free, open source computational fluid dynamics software. OpenFOAM is used across numerous engineering and science organizations, most notably in automotive, energy and aerospace. It’s designed to solve a wide range of problems, from complex fluid flows involving chemical reactions, turbulence and heat transfer, to acoustics, solid mechanics and electromagnetics. OpenFOAM takes advantage of the AMD EPYC processor’s ample memory bandwidth and large memory capacity.

 

For finite element analysis (FEA), AMD collaborated with LSTC. LS-DYNA® is a general-purpose multi-physics, finite element analysis program capable of simulating complex real-world problems. Widely used by the automotive industry to analyze vehicle designs, LS-DYNA® can accurately predict a car's behavior in a collision and the effects of the collision upon the car's occupants. These workloads are complex requiring a balance between floating point performance, memory bandwidth and network bandwidth. AMD EPYC processor’s eight lanes of memory bandwidth enable the system to more efficiently use the cores in each server. With LS-DYNA® and AMD EPYC processors, automotive companies and their suppliers can test car designs without having to tool or experimentally test a prototype, thus saving time and expense.

 

In addition, AMD is investing heavily in high-performance computing for weather related codes. WRF, IFS and HYCOM are all sophisticated applications used in research and operational forecasting. All require a balance of computational power, large volume data ingestion and memory bandwidth. Initial testing of AMD EPYC processor-based systems by the HPC and AI Innovation Lab showed impressive results on memory bandwidth and core density per socket making AMD EPYC processor-based servers a good choice for many applications. AMD is continuing to collaborate with the community to optimize the entire stack for all weather-related codes.

 

AMD is committed to continually expanding our partner ecosystem to create jointly engineered, optimized solutions for our customers that lower implementation risk and improve total cost of ownership.

 

Raghu Nambiar is the CVP & CTO of Datacenter Ecosystems & Application Engineering at AMD. His postings are his own opinions and may not represent AMD’s positions, strategies or opinions. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied.  GD-5