Run WRF from your browser
Because Cloudrun is web-based, you can use it to run WRF from anywhere.
Minimal set up required
You provide the input data and namelist configuration file, and we provide the hardware and WRF expertise. You don't have to worry about compilers, system libraries, or compute resources. Cloudrun brings compiled, and fully optimized WRF builds that take advantage of the latest hardware.
Pay only for what you use
We bill you on a per-run basis and only for the computational resources that you use.
All our WRF simulations run in parallel on dedicated servers with latest generation Intel Xeon Platinum 8168 processors. Our WRF builds have been fine-tuned to bring optimal performance.
Run any number of simulations concurrently
Cloudrun employs a swarm of scalable and parallel compute nodes. Unlike shared supercomputers, we don't limit how many runs you can make at a time.
Unlike on shared supercomputers, with Cloudrun you never need to wait for your runs to start. From uploading the input files to our platform, it typically takes a minute for your run to start.
Selection of run times and prices
When setting up a new run, Cloudrun parses your configuration file (
namelist.input) to determine the estimated completion time for different number of parallel compute settings. Runs that use fewer processors take longer to complete, but cost less. Cloudrun allows you to choose a run option that best fits your schedule and budget!
Cloud hosting of model output
With Cloudrun, you can choose to store your model output files in the cloud, or download them to your local machine. Our upcoming Cloudrun API will allow you to subset and slice model fields from your favorite programming language, without having to download the files.
WRF versions 3.8.1, 3.9, and 3.9.1
Currently, you can run only WRF 3.8.1 through 3.9.1, with the ARW dynamical core in real case mode (
em_real). We will support new versions of WRF as they come out. If you would like to run other models through Cloudrun, let us know!
Finally, an unparalleled WRF support
Have an issue setting up and configuring your WRF run? We got your back! Cloudrun team has a decade-long experience working with and contributing to the WRF codebase. No more you will have to scour through mailing lists and abandoned forums to find answers. We aim to be your go-to reference for any WRF questions.