Downloads 298this week

GitHub stars 12

Extension repository on GitHub

Extension descriptor (YAML)

GPU-accelerated analytical operators for DuckDB on NVIDIA CUDA and Apple Silicon Metal. First SQL execution engine that targets Apple Silicon GPUs.

Installing and Loading

INSTALL gpudb FROM community;
LOAD gpudb;

Example

LOAD gpudb;
-- Drop-in streaming aggregates (BIGINT and DOUBLE overloads)
SELECT gpu_sum(value::BIGINT) FROM range(1000000) AS t(value);

-- v0.4.0: resident columns — upload once, then reductions run on the GPU
-- with zero per-query transfer
SELECT gpu_upload('v', value::BIGINT) FROM range(1000000) AS t(value);
SELECT gpu_sum_resident('v');
SELECT gpu_last_stats();   -- which backend ran + kernel time

About gpudb

gpudb has two SQL surfaces:

Streaming aggregates — drop-in gpu_sum / gpu_min / gpu_max (BIGINT and DOUBLE overloads; smaller integers widen implicitly). They work in plain aggregation, GROUP BY, and window frames, match native DuckDB semantics (SQL NULL for empty/all-NULL input, NaN-aware total order for DOUBLE min/max), and run at parity with native — by design, since per-query GPU round-trips lose through this interface.

Resident columns (v0.4.0) — the path where the GPU wins. Upload a column to device memory once, then reductions run on-GPU with zero per-query transfer:

• gpu_upload(name, col) — one-time upload (BIGINT or DOUBLE)
• gpu_sum_resident / gpu_min_resident / gpu_max_resident (BIGINT)
• gpu_sum_resident_f64 (DOUBLE)
• gpu_resident_info, gpu_last_stats, gpu_build_info, gpu_drop_resident

Measured on TPC-H lineitem (DuckDB v1.5.5, 5-run medians, results verified equal to native; full grid with reproduction steps in the project's append-only BENCHMARK.md): SF50 SUM 99 ms native vs 4 ms resident on an RTX 4090 Laptop (25×, kernel at 563 GB/s ≈ VRAM bandwidth); SF100 (600M rows) 99 ms vs 10 ms on an Apple M4 Max (9.9×, Metal kernel at 503 GB/s). The speedup grows with data size. Honest trade-offs are documented alongside: whole-column min/max on stored tables stays a native win (zonemap statistics), one-shot cold queries favor native, and the one-time upload breaks even after roughly 100 repeated queries.

Community binaries ship the full Metal path on Apple Silicon and a clean CPU fallback on Linux (same SQL surface either way — LOAD never fails on GPU-less machines). The Linux build is CUDA-ready: it auto-enables the CUDA backend with a statically linked runtime when built with the CUDA toolchain, and SELECT gpu_build_info(); reports which backends a given binary carries.

Source: https://github.com/singhpratech/duckdbgpumetaldbram

Added Functions

Overloaded Functions

This extension does not add any function overloads.

Added Types

This extension does not add any types.

Added Settings

This extension does not add any settings.