PyFastFlow

⚠️ Experimental. This describes the in-progress v1 core. The API is settling and will change.

GPU routines for Earth-surface processes — portable across backends.

Two things in one:

• A portable GPU-routine engine. Compose parameters, helpers and multi-kernel routines once, and run them on Taichi, Quadrants, or CuPy — same composition model, backend-native kernels.
• A geomorphology toolbox built on it. Flow routing, flooding, and landscape evolution, engineered for grids of hundreds of millions of nodes.

Why

Writing flexible GPU kernels for physics simulation can be time-consuming. Performance on GPU is highly affected by memory layout. Hence simple things such as a constant vs a 2D varying parameter, or stencil operations (periodic boundaries, no-data, outlets, ...), each push you toward a different hand-tuned kernel — multiplied by every backend you want to run on. PyFastFlow lets you write a routine once and bind how each parameter and stencil is realised, then specialises it per backend and layout.

Example

import taichi as ti; ti.init(arch=ti.gpu)
from pyfastflow.experimental.core.context.taichi_backend import (
    TaichiKernelBuilder, TaichiParameter,
)
from pyfastflow.experimental.core.pool.taichi_pool import TaichiPool

pool = TaichiPool()
D = TaichiParameter("D", ti.f32, mode="scalar", value=1e-2, pool=pool)

def diffuse(z_out: ti.template(), z_in: ti.template()):
    for i, j in z_in:
        z_out[i, j] = z_in[i, j] + D.get(0) * (
            z_in[i-1, j] + z_in[i+1, j] + z_in[i, j-1] + z_in[i, j+1] - 4.0 * z_in[i, j])

step = TaichiKernelBuilder().bind("D", D).ingest(diffuse).compile()

The parameter / helper / routine API is identical on every backend; only the kernel body changes — a Python def on Taichi and Quadrants, CUDA source on CuPy. The same landscape-evolution model, built once per backend, lives in examples/core/lem/.

Install

pip install -e .            # from source

Requires Python ≥ 3.9, and the backend(s) you target (Taichi, Quadrants, CuPy).

Status

Core interface under active refactor toward v1.0 (+ paper). Flow/flood routines are being ported onto the new core; landscape evolution is WIP.

License & authors

CeCILL v2.1 — Boris Gailleton (Géosciences Rennes) · Guillaume Cordonnier (INRIA).