GitXplorerGitXplorer
g

swirl-c

public
1 stars
0 forks
0 issues

Commits

List of commits on branch main.
Unverified
43dc8fc8c0a62241d567f6fc9d6f30bc24672920

Removed redundant source code.

jjohn-qingwang committed 5 months ago
Unverified
6580de05187fdbc4c759d50051340b19711b1f6e

Removed all tests that dependends on Google internal libraries.

jjohn-qingwang committed 5 months ago
Unverified
f3a11de8c0508509f6e0b7ceb853071c7dd0caab

Remove all BUILD files.

jjohn-qingwang committed 5 months ago
Unverified
e643b81e307adb75ef4c8a13acebfad79ca2ab84

Added license headers to all source code file.

jjohn-qingwang committed 5 months ago
Unverified
edfc00ad55be11635a48278a00ae876ae198a189

Remove unpublished files.

jjohn-qingwang committed 5 months ago
Unverified
81a60b0910f63643fe6fe9aeedd65165e9e53ebe

Adding source code for Swirl-C version 0.0.

jjohn-qingwang committed 5 months ago

README

The README file for this repository.

Swirl-C: Computational Fluid Dynamics in TensorFlow (Compressible Flows)

This is not an official Google product

Swirl-C is a computational fluid dynamics (CFD) simulation framework that is accelerated by the Tensor Processing Unit (TPU) for compressible flows. It solves the three dimensional Navier-Stokes equation with a fully compressible approach, and the governing equations are discretized by a finite-volume method on a collocated structured mesh. It is implemented in TensorFlow.

Installation

To use Swirl-C, you will need access to TPUs on Google Cloud. For small simulations, the easiest way to access TPUs is to use Google Colab. To run large simulations, you will need to create TPU Nodes or VMs in your Google Cloud project.