GPU Setup¶

OpenBoost uses CUDA for histogram building and tree construction. NaturalBoost, FormulaBoost, and WeibullAFT share that tree path. Some objective math still runs on the host (FormulaBoost GGN today; LogNormal / digamma families). Trees are the expensive part at scale.

Verify detection¶

import openboost as ob

print(ob.get_backend())   # "cuda" or "cpu"
print(ob.is_cuda())       # True if a GPU is active

Install the extra first: pip install --pre "openboost[cuda]".

Pin a backend¶

import openboost as ob

ob.set_backend("cpu")    # debug / comparison
ob.set_backend("cuda")

# Or:
# export OPENBOOST_BACKEND=cuda

backend_context("cpu") is a temporary switch that restores the previous backend on exit.

What the A100 numbers actually are¶

NaturalBoost vs NGBoost, heteroscedastic Normal, 80 features, 500 trees, Modal A100. NGBoost has no GPU implementation, so this is GPU OpenBoost against CPU NGBoost, which is the comparison that exists in the world.

n_train	OpenBoost (A100)	NGBoost (CPU)	speedup
45K	3.01s	1414s	470×
90K	2.21s	2716s	1229×
450K	5.40s	skipped (hours)	n/a
900K	6.94s	skipped	n/a

NLL is tied at every size that both ran. Full tables: Benchmarks.

When GPU helps

Histogram trees win at tens of thousands of rows and up. Below ~5K samples, kernel launch overhead often matches CPU. Use float32 features. Missing values and categoricals fall back from the GPU-native builder to the hybrid path (warning emitted).

On CPU, NaturalBoost and NGBoost are ~parity (0.8–1.3×). Do not quote the A100 ratio as a CPU claim.

FormulaBoost and WeibullAFT on GPU¶

Both use GPU trees. FormulaBoost's GGN (finite-difference Jacobian + K×K solve) currently runs on the host; at 200K rows full GGN still beat an XGBoost custom objective (17s vs 20s on A100). WeibullAFT's expected Fisher step is cheap (2×2 per row).

Multi-GPU¶

import openboost as ob

model = ob.GradientBoosting(n_trees=100, n_gpus=4)
model.fit(X, y)

model = ob.GradientBoosting(n_trees=100, devices=[0, 2])
model.fit(X, y)

Requires pip install --pre "openboost[distributed]" (Ray). NaturalBoost / FormulaBoost / WeibullAFT currently train on one GPU.

Requirements¶

NVIDIA GPU, CUDA Compute Capability 3.5+
CUDA Toolkit 11 or 12
numba-cuda>=0.23

Troubleshooting¶

Training seems slow on GPU. Features should be float32. Tiny datasets do not amortize kernel launch. Confirm ob.is_cuda() is True.

CUDA not detected.

nvidia-smi
python -c "from numba import cuda; print(list(cuda.gpus))"
Reinstall openboost[cuda] against the CUDA version on the machine

Trained on GPU, loading on CPU. Saved models are backend-agnostic.

model.save("model.joblib")
loaded = ob.NaturalBoostNormal.load("model.joblib")  # CPU or GPU