docs.rs failed to build skippydb-0.2.1
Please check the build logs for more information.
See Builds for ideas on how to fix a failed build, or Metadata for how to configure docs.rs builds.
If you believe this is docs.rs' fault, open an issue.

skippydb

A high-performance verifiable key-value store optimized for blockchain state storage, featuring SHA256 Merkle trees with optional CUDA GPU acceleration.

Installation

cargo add skippydb
# With GPU acceleration:
cargo add skippydb --features cuda

Quick Start

use skippydb::config::Config;
use skippydb::AdsCore;

let config = Config::from_dir("/path/to/data");
AdsCore::init_dir(&config);

Feature Flags

Feature	Description	Default
`default`	Includes jemalloc allocator	Yes
`cuda`	CUDA GPU-accelerated SHA256 batch hashing	No
`directio`	Linux io_uring direct I/O for reads	No
`hpfile_all_in_mem`	Keep heap files entirely in memory	No
`tee_cipher`	AES-GCM encryption for data at rest	No
`slow_hashing`	Slow hashing mode for testing	No

GPU Acceleration

With --features cuda, KyumDB batches Merkle tree SHA256 operations to CUDA cores:

3-5x throughput improvement at batch sizes >10K
Automatic kernel selection (CPU/AoS/SoA) via auto_batch_node_hash
Async pipelining, fused active bits, GPU-resident upper tree
Multi-GPU support via MultiGpuHasher

See the GPU Integration Guide for usage patterns.

Documentation

License

Licensed under either of:

Apache License, Version 2.0
MIT License

at your option.