Book Our ADCIRC Live Cluster

Our Services

Accessing our managed ADCIRC cluster is not free, but it's very cost effective. It is not running in the cloud, it is a physical cluster we run and maintain in Houston, TX, USA. We are still determining the best way to provide services, but currently we ask interested users to email us.At this time we are able to block out a day weeks of time where you may use the cluster exclusively, save the operational emergency (a hurricane), when we need to take it over. Exclusive access to our cluster starts at $100/day for a minimum of 3 days. We offer direct support and additional services that may increase that rate. Clients should come with friendly attitude :-).

Email us to book the full cluster, starting at $100/day

About Our Cluster

We manage an 80 CPU core cluster, and we are in the process of upgrading to 160 CPU cores. Rather than nickle and dime users with CPU/hrs, we wish to find out your needs and provide a custom quote. We are open to renting out part or all of the cluster for fixed blocks of time - based on timing and priority (we use the cluster, too!), and at costs far less than what would be available on commercial cloud services. Our goal is the same as yours, for you to get your runs done. Let us know!

Email us at help@support.adcirc.live to request information on renting our cluster and free quote!

Our ADCIRC Cluster Specifications

Component	Details
Compute Nodes	2 × Dell Precision 7810 Workstations (upgrading to 4 total)
Cores	80
Processors	Each node: 2 × Intel Xeon E5-2698 v4 (20 cores, 40 threads, 2.2 GHz base, 3.6 GHz turbo) Total per node: 40 physical cores Current cluster: 80 physical cores Planned upgrade: 160 physical cores
Memory	32 GB DDR4 RAM per node (64 GB total)
Storage	16 TB shared storage
Job Scheduler	Slurm workload manager
Software	Ubuntu Linux Prebuilt latest ADCIRC, SWAN available Intel OneAPI HPC Compiler Suite Docker (compute01)
Estimated Peak Performance	Intel Xeon E5-2698 v4 has a peak of ~0.88 TFLOPS (double precision) per CPU. 4 CPUs (2 per node × 2 nodes) → ~3.5 TFLOPS (DP theoretical peak). Single precision performance ~2× higher → ~7.0 TFLOPS. This cluster is well-suited for mid-scale to large coastal modeling studies.

ADCIRC Scaling Rule of Thumb:

ADCIRC scales at ~5,000 mesh nodes per physical core.
• Current capacity (80 cores): ~400,000 mesh nodes.
• Planned upgrade (160 cores): ~800,000 mesh nodes.

If you need a cost effective environment to scale you ADCIRC capacity quickly and cheaply, click here to request a free quote!

ADCIRC Cluster Services

There are many needs associated with running ADCIRC locally, and we are prepared to discuss them all with you. In particular:

Pros/Cons of a local cluster versus cloud, when to use either
Pros/Cons of multi-core workstations or laptops over a local cluster
Practical considerations: electical draw, addition HVAC requirements
Cost, total cost of ownership, depreciation of equipment; when to by used versus new, etc

We're also able to provide full service for your cluster, including:

Design and equipment listing
Cluster installation and set up
ADCIRC installation
ADCIRC cluster user training or IT support
ADCIRC engineering support
General cluster maintenance, batch user (Slurm) set up, etc
Rent our cluster

If this sounds like something you are interested in learning more about, click here to request more information or receive a free quote!

ASGS
CERA
Paraview
HDFView