Computational fluid dynamics (CFD) is a computer-aided engineering (CAE) tool used to simulate fluid flow around various engineering models. CFD applications are most efficiently run on high performance computing (HPC) systems with large throughput local storage, low latency networks, and optimized CPUs.
Oracle Cloud Infrastructure now offers HPC clusters with CFD software that can easily be deployed through Oracle Cloud Marketplace. The compute cluster is provisioned with the following:
OpenFOAM is installed with scripts to execute tutorials simulations.
Libraries required for some of the most commonly used commercial CFD applications, including Converge, StarCCM+ and Ansys Fluent, are installed. Just bring your own license to get your simulations up and running.
High throughput NFS and GlusterFS file servers are installed on NVMe SSD local storage to share data across the compute cluster and maximize data throughput.
Ansys Fluent Runbook
Oracle and Intel partnered with performance solutions company MV Concept to optimize Ansys Fluent on Oracle Cloud Infrastructure. By using Intel mpi’s and libraries, you can further maximize your performance with the BM.HPC2.36 Intel Skylake 6154 processor. Learn more about the recommended tips and tricks using Intel technology.
CFD Runbook
Learn how to launch a CFD-ready cluster and run some of the most common CFD applications on Oracle Cloud Infrastructure.
As CFD continues to grow in popularity, so does the need to run on robust and performant infrastructure. By moving or supplementing with Oracle Cloud Infrastructure, users have a variety of instances to choose from for their HPC workloads. These instances are some of the most high-performing in the cloud, and are supported in multiple regions across the globe at consistent pricing. Although any compute instance may be used for HPC purposes, below are some of the most popular and recommended for CFD.
Instance* |
Specifications |
Application |
Billing |
Technical |
---|---|---|---|---|
BM.HPC2.36 | Intel Skylake Processors 6.4 TB NVMe SSD storage 100 Gbps RDMA networking |
Geared towards applications that are optimized for CPUs only and can be run in parallel across a large number of cores. | Per second billing with a 1 hour minimum. No charge for RDMA or NVMe SSD local storage. | Optimize your HPC with Oracle Cloud on Intel |
BM.GPU4.8 | AMD Processors 8 x 200 Gbps RDMA 2048 GB memory 27.2 TB NVMe SSD storage |
Geared towards applications that are designed to run on GPUs. RDMA cluster networking is an added bonus for applications that can run in parallel. | Per second billing with a 1 hour minimum. No charge for RDMA or NVMe SSD local storage. | NVIDIA A100 on Oracle Cloud Infrastructure |
BM.GPU3.8 | 8 NVIDIA V100 GPU’s Intel Processors Block storage only 2 x 25 Gbps networking |
Geared towards applications that are designed to run on GPUs. | Per second billing with a 1 hour minimum. Block storage attachment is defined by the user and billed separately. | Rocky on Oracle Cloud |
BM.Standard.E3.128 | AMD Processors 2048 GB memory Block storage only 2 x 50 Gbps networking |
Geared towards high-memory workloads that do not need to scale past a couple of instances. | Per second billing with a 1 hour minimum. Block storage attachment is defined by the user and billed separately. | Oracle Cloud Infrastructure Compute E3 Platform on 2nd Gen AMD EPYC Processors |
VM.Standard.E3.Flex | AMD Processors Flexible memory and CPUs Block storage only Max 40 Gbps networking |
Geared towards workloads that are cost-sensitive but not performance-sensitive. | Per second billing with no minimum. Block storage attachment is defined by the user and billed separately. | HPC on VM.Flex |
*Additional instances suited for CFD are planned for release in 2021:
Innovate the Future of Medical Treatment through Virtual Humans
Kristen Yang, Senior Technical Program Manager, Oracle
Oracle continues to push the limits in the cloud on both performance and cost. By running high-performance computing (HPC) on Oracle Cloud Infrastructure, you can get the same performance as on-premises, with the added benefits of being able to deploy instantly and eliminating hardware acquisition and maintenance costs. This opportunity has influenced many enterprise customers to move their HPC workloads to Oracle Cloud. For examples, see Nissan Moves to Oracle Cloud and Oracle Cloud Infrastructure Behind World’s Fastest Supercomputer. The types of workloads running on Oracle Cloud extend into every industry, with some of the most common ones being computational fluid dynamics (CFD) and crash simulation applications.
Read the complete postBuild, test, and deploy applications on Oracle Cloud—for free. Sign up once, get access to two free offers.
Explore cloud training resources with Oracle Cloud Infrastructure training videos, self-paced learning labs, and certifications.
Oracle Cloud Infrastructure combines the elasticity and utility of the public cloud with the control, security, performance, and predictability of on-premises computing environments.