PoSpace Reference Implementation (CHIP)

This repository provides a public reference implementation of the new Proof of Space (“PoSpace”) format, as defined in CHIP-48. It includes:

• lib/pos/ — core Proof of Space header‑only library (<pos/...>)
• lib/common/ — header‑only utilities (<common/...>)
• lib/fse/ — vendored FSE compression library
• tools/src/plotter/ — example C++ plotter executable using the above
• tools/src/solver/ - solver benchmarking and testing

---

Features

• Header‑only, C++20 PoSpace core (ProofCore, hashing, parameters, validator, etc.)
• Example plotter tool demonstrating the plot pipeline and writing a plot (but does not yet compress data)
• Solver tool for benchmarking solve times for k28/30/32 sizes on CPUs.

Note

The reference implementations do not optimize for memory usage, and consume much more memory than the upcoming production versions.

---

Prerequisites

• A C++20‑capable compiler
• CMake ≥ 3.15
• make (or your preferred build tool)
• A Unix‑style shell (Linux/macOS) or PowerShell/Bash on Windows

Building

1. Clone the repo:

   git clone https://github.com/Chia-Network/pos2-chip.git

   cd pos2-chip

2. Build

   Option A: Use the helper script

   ./build-release.sh

   Option B: Use CMake

   First, build with Release mode to enable optimizations:

   cmake -B build -DCMAKE_BUILD_TYPE=Release .

   Next, compile:

   Linux:

   cmake --build build -j$(nproc)

   macOS:

   cmake --build build -j$(sysctl -n hw.logicalcpu)

---

Running the Plotter

From the root of your build directory, invoke the plotter executable:

./build/src/tools/plotter/plotter <k> [sub_k]

By default it uses the sample plot ID and parameters defined in tools/plotter/src/main.cpp. To customize, edit that file or supply your own main() implementation.

About k

k determines the size of the plot. For testing, valid k-sizes are even numbers from 18 to 32. For mainnet, only 28, 30, and 32 will be allowed.

For your reference, the current plot size for k=18 is ~6.5 MB. For k=28, it's ~5 GB. The final optimized sizes are expected to be about 75% smaller than the current sizes.

About sub_k

Each increment of sub_k doubles the number of unique Proof Fragments an attacker must solve for. The goal is to achieve bit drop saturation in order to maximize resistance against compression.

The only k/sub_k combinations expected to be valid on mainnet are:

• k=28, sub_k=20
• k=30, sub_k=21
• k=32, sub_k=22

If running code-based or CI tests, then k=18, sub_k=15 is recommended.

For running on a testnet, k=24, sub_k=18 will likely be recommended. The reason for the larger values on testnet is to lower the variance on plot size and output.

Examples

To use k=18 with the default sub_k:

./build/src/tools/plotter/plotter 18

To use k=28 and sub_k=20

./build/src/tools/plotter/plotter 28 20

Running the Solver

Run the solver executable with one of the two modes:

Coming soon.

Benchmark Mode

./build/src/tools/solver/solver benchmark <k-size>

Where <k-size> is an integer value for the solver’s k parameter (e.g. 28).

Example:

./build/src/tools/solver/solver benchmark 32

Outputs timing and performance metrics for reconstructing proofs.

Prove Mode

Reads the plot, prints its parameters, and runs the a chaining test for getting and solving for a full proof.

Note

Currently the solver does not accept a challenge to choose a proof from the plot. Coming soon (TM).

./build/src/tools/solver/solver prove <plot-file>

Where <plot-file> is the path to a plot file to test.

Example:

./build/src/tools/solver/solver prove /path/to/plot.bin

Note

Plot files are changing frequently, so use the plotter to generate a new plot to then test it.