mirror of
https://github.com/hanzoai/benchmarks.git
synced 2026-10-30 14:59:11 +00:00
zeekay
a93a975b61
Real-world benchmarks comparing ZAP vs JSON/Protobuf: Serialization (Go): - Tool call encode: 21x faster - Passthrough routing: 656x faster - Batch 100 messages: 21x faster, 0 allocations Blockchain/Consensus (Go): - Warp message: 70x faster encode, 244x faster decode - Consensus vote: 1,438x faster - State access: 736,458x faster (mmap simulation) MCP/Agent (Python): - Memory: 99.7% reduction (825MB → 2.4MB for 100 servers) - 20-agent orchestration: 2x faster, 8x less memory Distributed Inference (Python): - KV cache transfer: 926x faster - Batch prompts: 200x faster, 68% smaller - Speculative decode: 210x faster Benchmarks for: - Warp messaging (cross-chain) - Consensus voting and attestations - Multi-agent tool calling (MCP) - Distributed AI inference
5 lines
61 B
Plaintext
5 lines
61 B
Plaintext
psutil>=5.9.0
|
|
numpy>=1.24.0
|
|
matplotlib>=3.7.0
|
|
aiohttp>=3.9.0
|