vibespatial 0.3.0

GPU-first spatial analytics for Python — drop-in GeoPandas replacement backed by CUDA kernels

vibeSpatial

vibeSpatial is a GPU-first spatial analytics library for Python. Change one import line and your existing GeoPandas code runs on CUDA — binary predicates, buffer, overlay, dissolve, make-valid, spatial joins, and I/O all dispatch to GPU kernels automatically, with explicit, observable CPU compatibility fallback only when the native GPU path is unavailable or unsupported.

[!WARNING] vibeSpatial is very early in development. Operations may be unoptimized or have multiple Host/Device transfers causing reduced performance. File an issue if you hit a problem!

The repository enforces fallback observability: once a workflow is on device, hidden host exits are treated as bugs, and strict-native tests fail if a path materializes to host without first recording an explicit fallback or compatibility boundary. The maintained warmed 10k shootout suite under benchmarks/shootout/ is currently at or above parity on local RTX 4090 runs.

Install

pip install vibespatial              # CPU-only (GeoPandas drop-in)
pip install vibespatial[cu12]        # CUDA 12 GPU acceleration
pip install vibespatial[cu13]        # CUDA 13 GPU acceleration

Quick start

import vibespatial as gpd

gdf = gpd.read_file("my_data.gpkg")
buffered = gdf.buffer(100)
joined = gpd.sjoin(gdf, buffered)
gdf.to_parquet("out.parquet")

Real-world example: 7.2 million buildings

Load every building footprint in Florida, reproject to UTM, find all buildings within 1 km of a random pick, and export to GeoParquet. The full script is at examples/nearby_buildings.py.

import vibespatial as gpd

# Read 7.2M buildings from Microsoft US Building Footprints
gdf = gpd.read_file("Florida.geojson")

# Reproject to UTM for metric distances
gdf_utm = gdf.to_crs(gdf.geometry.estimate_utm_crs())

# Pick a random building and find everything within 1 km
seed = gdf_utm.geometry.iloc[random.randrange(len(gdf_utm))]
nearby = gdf_utm[gdf_utm.geometry.dwithin(seed.centroid, 1_000)]

# Export to GeoParquet
nearby.to_crs(epsg=4326).to_parquet("nearby_buildings.parquet")

vibeSpatial is a drop-in replacement for GeoPandas. Here is the only diff:

-import geopandas as gpd
+import vibespatial as gpd

 gdf = gpd.read_file("Florida.geojson")
 gdf_utm = gdf.to_crs(gdf.geometry.estimate_utm_crs())
 seed = gdf_utm.geometry.iloc[random.randrange(len(gdf_utm))]
 nearby = gdf_utm[gdf_utm.geometry.dwithin(seed.centroid, 1_000)]
 nearby.to_crs(epsg=4326).to_parquet("nearby_buildings.parquet")

Performance on 7.2M polygons (RTX 4090 vs GeoPandas on i9-13900k):

Step	GeoPandas	vibeSpatial	Speedup
Read GeoJSON	57.7 s	11.7 s	4.9x
Reproject to UTM	8.2 s	0.4 s	21x
Select within 1 km	0.2 s	0.5 s	--
Write GeoParquet	0.2 s	0.1 s	2x
End-to-end	66.3 s	12.7 s	5.2x

GeoJSON reading uses GPU byte-classification: 10 NVRTC kernels parse JSON structure, extract coordinates, and assemble geometry directly on-device in 1.8 s (32x vs pyogrio); property extraction stays on CPU via orjson. Reprojection uses vibeProj fused GPU kernels via transform_buffers() -- no host round-trip. Spatial queries use device-resident bounding-box prefilter + GPU distance kernels.

Tech stack

Layer	Technology
GPU kernels	NVRTC (runtime-compiled CUDA C via cuda-python)
GPU primitives	CCCL (cccl — scan, sort, reduce, select)
GPU arrays	CuPy (device memory, element-wise ops, prefix sums)
GPU JSON parse	Custom byte-classification kernels (ADR-0038)
GPU projection	vibeProj
GPU Parquet/Arrow	pylibcudf (WKB decode, GeoArrow codec)
CPU compatibility	GeoPandas API (vendored upstream test suite)
JSON parsing	orjson (property extraction)
File I/O	pyogrio (Shapefile, GPKG, small GeoJSON)
Packaging	uv, hatchling

All GPU kernels are pure Python — CUDA C source strings compiled at runtime via NVRTC with background warmup (ADR-0034). Compiled CUBINs are cached on disk so the JIT cost is paid only once per install. No compiled extensions, no nvcc build step. The entire suite ships as pure-Python wheels:

Package	Wheel size
vibespatial	612 KB
vibeproj	57 KB
vibespatial-raster	51 KB
Total	720 KB

Pre-compilation

The first time a GPU operation runs, CUDA kernels are JIT-compiled in the background (~2-3 s wall time on 8 threads). Compiled CUBINs are cached on disk so subsequent process starts are near-instant. To pre-populate the caches (e.g. in CI or after install):

from vibespatial.cccl_precompile import precompile_all
precompile_all()  # compiles all 21 CCCL specs + 61 NVRTC kernels, blocks until done

Or from the command line:

uv run python -c "from vibespatial.cccl_precompile import precompile_all; precompile_all()"

See GPU Kernel Caching for the full design and environment variables.

Documentation

See the documentation for the full API reference, GPU acceleration guide, and I/O format support matrix.

---

Contributing

uv sync --group dev
uv run python scripts/check_docs.py --refresh
uv run python scripts/vendor_geopandas_tests.py
uv run pytest tests/upstream/geopandas/tests/test_config.py

Dependency groups

• dev: local development and pytest tooling
• upstream-optional: heavier I/O and visualization extras for broader coverage
• gpu-optional: CUDA runtime, CuPy, pylibcudf

Layout

• src/vibespatial/: package code
• src/geopandas/: GeoPandas compatibility shim
• tests/: repo-owned tests
• tests/upstream/geopandas/: vendored upstream GeoPandas test suite
• docs/: architecture docs and ADRs
• examples/: benchmarks and usage examples