You can even mount a DuckLake - meaning you can have a "data lake in your pocket".
It is fully secure, and runs the TPC-H 1GB 22-query benchmark in less than 2 seconds on an iPhone 16 Pro Max.
See: https://github.com/gizmodata/gizmosql for more details...
Instance: r8gd.metal-48xl (Graviton4) 192 vCPUs, 1536 GiB RAM 11.4TB RAID0 NVMe on AWS $14.11 / hr on-demand $2.82 / hr spot
Dataset: 2.3TB Parquet files (100,000 files with 10,000,000 rows each!) Copy from S3: 11m 24s Challenge Query (cold start): 2m 22s Challenge Query (warm start): 2m 09s Total cost: ~$0.11 (spot) per query And yes...
SELECT COUNT(*) on 1 trillion rows? Just 21.8 seconds.
Why it matters: GizmoSQL lets you query massive Parquet datasets interactively, directly, and efficiently – using the DuckDB SQL engine, and the Arrow Flight SQL protocol for blazing-fast client connections.
No complex Spark or distributed compute cluster setup. No hidden egress fees. Just pure, fast SQL on massive data.
Check out open-source GizmoSQL here: https://github.com/gizmodata/gizmosql and https://gizmosql.com
GizmoSQL is free for self-hosting!
Full 1TRC details, scripts, and results: https://github.com/coiled/1trc/issues/7
I'm Philip Moore - the founder of GizmoData, and creator of GizmoEdge - a Distributed SQL Engine powered by Internet-of-Things (IoT) devices.
GizmoEdge is a prototype application - that allows you to run SQL queries that get distributed to many devices - including: Linux, macOS, iOS, iPadOS, Kubernetes Pods, Raspberry Pis, and more.
I've built a front-end application from which you can issue distributed SQL queries - here: https://gizmoedge.gizmodata.com
If you have an Apple device - you can install the GizmoEdge Worker app to have it "join the collective" - here is the App Store link: https://apps.apple.com/us/app/gizmoedge/id6738658135
Once you install the app, you just connect it up to the running GizmoEdge server - and it will download a shard of TPC-H data at the 1GB foot-print (it is actually smaller, b/c it is compressed parquet - in a ZStandard-compressed tarball). The app will then create a DuckDB database from the parquet datasets that are on your device. Right now - queries will only run on your device if the app is open and in the foreground. To connect your device to the server - click the little blue "server" icon just to the right of the GizmoData logo at the top of the screen - the username, password, etc. are pre-filled for you - just click the blue: "Connect WebSocket" button at the bottom to get started.
When you issue a SQL query that is eligible for distribution from the GizmoEdge SQL app (at https://gizmoedge.gizmodata.com) - it should distribute and execute on your device (among others) in the collective, provided that they have downloaded a shard of data and are connected to the server.
I would be honored if folks tried the engine. Please bear in mind - this is a prototype - it isn't production ready, yet. It is currently read-only as well - but I'm working on a way to make it easy for folks to ingest data for distributed SQL execution in the near-future.
GizmoEdge is powered by DuckDB on the workers, and uses WebSockets for low-latency connections between the server and workers. It uses TLS for encryption of communication - and has a robust security model - in which the server and workers have a "trust-but-verify" relationship.
Getting started: GizmoEdge SQL Navigator app (run interactive SQL queries here): https://gizmoedge.gizmodata.com GizmoEdge Worker on the App Store: https://apps.apple.com/us/app/gizmoedge/id6738658135 GizmoEdge homepage: https://gizmodata.com/gizmoedge
Thank you for your time and feedback - in advance!
