BespokeOLAP Artifacts

Generated C++ artifacts for Bespoke OLAP: synthesized implementations of the TPC-H and CEB benchmarks.

Setup

Requires uv.

uv sync

Build & Run Generated Databases

We have provided a sample script to run the generated databases in run.py.

python run.py [--sf <scale_factor>] [--no-optimize] {tpch,ceb}

Arguments:

Argument	Description	Default
benchmark	Benchmark to run: tpch or ceb	(required)
--sf	Scale factor	1
--no-optimize	Disable compiler optimization	(optimization on)

Examples:

# Run TPC-H at scale factor 20
python run.py --sf 20 tpch

# Run CEB at scale factor 1
python run.py ceb

# Run TPC-H without optimization flags (compiler)
python run.py --no-optimize --sf 10 tpch

Benchmarks

• bespoke_tpch/ — TPC-H queries Q1–Q22, synthesized as bespoke C++ implementations
• bespoke_ceb/ — CEB queries, Q1a-Q11b synthesized as bespoke C++ implementations