{"id":47787971,"url":"https://github.com/Lattice9AI/NexusFix","last_synced_at":"2026-04-08T21:01:06.847Z","repository":{"id":335018921,"uuid":"1139441042","full_name":"Lattice9AI/NexusFix","owner":"Lattice9AI","description":"A zero-alloc, compile-time hardened FIX engine built for sub-100ns execution.","archived":false,"fork":false,"pushed_at":"2026-03-31T23:42:31.000Z","size":9209,"stargazers_count":42,"open_issues_count":1,"forks_count":7,"subscribers_count":4,"default_branch":"main","last_synced_at":"2026-04-01T01:53:15.591Z","etag":null,"topics":["cpp23","fix-protocol","header-only","hft","high-frequency-trading","low-latency","quantitative-finance","simd","trading"],"latest_commit_sha":null,"homepage":null,"language":"C++","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Lattice9AI.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":"CODE_OF_CONDUCT.md","threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":"SECURITY.md","support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-01-22T00:51:10.000Z","updated_at":"2026-03-31T23:42:34.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/Lattice9AI/NexusFix","commit_stats":null,"previous_names":["silverstreamsai/nexusfix","lattice9ai/nexusfix"],"tags_count":16,"template":false,"template_full_name":null,"purl":"pkg:github/Lattice9AI/NexusFix","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Lattice9AI%2FNexusFix","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Lattice9AI%2FNexusFix/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Lattice9AI%2FNexusFix/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Lattice9AI%2FNexusFix/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Lattice9AI","download_url":"https://codeload.github.com/Lattice9AI/NexusFix/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Lattice9AI%2FNexusFix/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31573788,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-08T14:31:17.711Z","status":"ssl_error","status_checked_at":"2026-04-08T14:31:17.202Z","response_time":54,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.5:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["cpp23","fix-protocol","header-only","hft","high-frequency-trading","low-latency","quantitative-finance","simd","trading"],"created_at":"2026-04-03T15:05:03.627Z","updated_at":"2026-04-08T21:01:06.829Z","avatar_url":"https://github.com/Lattice9AI.png","language":"C++","readme":"\u003cp align=\"center\"\u003e\n  \u003ch1 align=\"center\"\u003eNexusFIX\u003c/h1\u003e\n  \u003cp align=\"center\"\u003e\n    \u003cstrong\u003eUltra-Low Latency FIX Protocol Engine for High-Frequency Trading\u003c/strong\u003e\n  \u003c/p\u003e\n  \u003cp align=\"center\"\u003e\n    Modern C++23 | Zero-Copy | SIMD-Accelerated | 3x Faster than QuickFIX\n  \u003c/p\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"https://stratcraft.ai/nexusfix\"\u003e\u003cimg src=\"https://img.shields.io/badge/Website-stratcraft.ai-blue.svg\" alt=\"Website\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://github.com/Lattice9AI/NexusFix/actions/workflows/ci.yml\"\u003e\u003cimg src=\"https://github.com/Lattice9AI/NexusFix/actions/workflows/ci.yml/badge.svg\" alt=\"CI\"\u003e\u003c/a\u003e\n  \u003cimg src=\"https://img.shields.io/badge/C%2B%2B-23-blue.svg\" alt=\"C++23\"\u003e\n  \u003cimg src=\"https://img.shields.io/badge/Library-Header--Only-blue.svg\" alt=\"Header-Only\"\u003e\n  \u003cimg src=\"https://img.shields.io/badge/License-MIT-green.svg\" alt=\"MIT License\"\u003e\n  \u003cimg src=\"https://img.shields.io/badge/Platform-Linux%20|%20macOS%20|%20Windows-lightgrey.svg\" alt=\"Cross-Platform\"\u003e\n  \u003cimg src=\"https://img.shields.io/badge/Speed-3x%20QuickFIX-orange.svg\" alt=\"3x Faster\"\u003e\n  \u003cimg src=\"https://img.shields.io/badge/Allocations-Zero%20on%20Hot%20Path-brightgreen.svg\" alt=\"Zero Allocation\"\u003e\n  \u003ca href=\"https://github.com/wilsonfreitas/awesome-quant\"\u003e\u003cimg src=\"https://awesome.re/mentioned-badge.svg\" alt=\"Mentioned in Awesome\"\u003e\u003c/a\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"#performance\"\u003ePerformance\u003c/a\u003e |\n  \u003ca href=\"#architecture-influences\"\u003eArchitecture\u003c/a\u003e |\n  \u003ca href=\"#features\"\u003eFeatures\u003c/a\u003e |\n  \u003ca href=\"#quick-start\"\u003eQuick Start\u003c/a\u003e |\n  \u003ca href=\"#documentation\"\u003eDocs\u003c/a\u003e |\n  \u003ca href=\"#contact\"\u003eContact\u003c/a\u003e\n\u003c/p\u003e\n\n---\n\n## Why NexusFIX?\n\n**NexusFIX** is a high-performance **FIX protocol** (Financial Information eXchange) engine built for **ultra-low latency quantitative trading**, **sub-microsecond algorithmic execution**, and **high-frequency trading (HFT)** systems. It solves the **performance bottlenecks** of traditional FIX engines by utilizing **hardware-aware C++ programming**.\n\nNexusFIX serves as a modern, faster alternative to QuickFIX with **zero heap allocations** on the critical path.\n\n\u003e *\"If you're building a low-latency trading system and QuickFIX is your bottleneck, NexusFIX is your solution.\"*\n\n---\n\n## Performance\n\n### NexusFIX vs QuickFIX Benchmark\n\nTested on Linux with GCC 13.3, 100,000 iterations:\n\n| Metric | QuickFIX | NexusFIX | Improvement |\n|--------|----------|----------|-------------|\n| **ExecutionReport Parse** | 730 ns | 246 ns | **3.0x faster** |\n| **NewOrderSingle Parse** | 661 ns | 229 ns | **2.9x faster** |\n| **Field Lookup** (O(1) post-parse, 4 fields) | 31 ns | 11 ns | **2.9x faster** |\n| **Parse Throughput** | 1.19M msg/sec | 4.17M msg/sec | **3.5x higher** |\n| **P99 Parse Latency** | 784 ns | 258 ns | **3.0x lower** |\n\n### Why is NexusFIX Faster?\n\n| Technique | QuickFIX | NexusFIX |\n|-----------|----------|----------|\n| Memory | Heap allocation per message | Zero-copy `std::span` views |\n| Field Lookup | O(log n) `std::map` | O(1) direct array indexing |\n| Parsing | Byte-by-byte scanning | AVX2 SIMD vectorized |\n| Field Offsets | Runtime calculation | `consteval` compile-time |\n| Enum/Type Conversion | Runtime switch chains (~300 branches) | 22 compile-time lookup tables (55-97% faster) |\n| Error Handling | Exceptions | `std::expected` (no throw) |\n\n### Zero Allocation Proof\n\nParsing a **NewOrderSingle** message on the hot path:\n\n| Operation | QuickFIX | NexusFIX |\n|-----------|----------|----------|\n| **Heap Allocations** | ~12 (`std::string`, `std::map` nodes) | **0** |\n| **Field Storage** | `std::map\u003cint, std::string\u003e` copies | `std::span` views into original buffer |\n| **Parsing Logic** | Runtime map insertion | Compile-time offset table |\n| **Memory Footprint** | Dynamic, unpredictable | Static, pre-allocated PMR pool |\n| **Destructor Overhead** | ~12 `std::string` destructors | **0** (no owned memory) |\n\n*Verified via custom allocator instrumentation. See [Optimization Diary](docs/optimization_diary.md).*\n\n*For kernel bypass (DPDK/AF_XDP) and FPGA acceleration, see [Roadmap](docs/design/TICKET_204_AERON_HIGH_THROUGHPUT_MESSAGING.md).*\n\n---\n\n## Architecture Influences\n\nNexusFIX stands on the shoulders of giants. We systematically studied **11 industry-leading Modern C++ libraries** and applied their techniques to ultra-low latency FIX processing. Below is our learning journey: what we learned, what we built, and what improvement we measured.\n\n### Learning → Implementation → Verification\n\n| Source Library | Engineering Evaluation | What We Changed | Benchmark Result |\n|----------------|------------------------|-----------------|------------------|\n| [hffix](https://github.com/jamesdbrock/hffix) | O(n) iterator-based field lookup is suboptimal for dense FIX packets; lacks compile-time optimization and type safety | `[Optimized]` `consteval` field offsets + `std::span` zero-copy views + O(1) direct indexing | **14ns** field access vs ~50ns iterator scan |\n| [Abseil](https://github.com/abseil/abseil-cpp) | Swiss Tables offer SIMD-accelerated probing with 7-bit H2 fingerprints; superior cache locality for session maps | `[Adopted]` `absl::flat_hash_map` for session store | **[31% faster](docs/compare/ABSEIL_FLAT_HASH_MAP_BENCHMARK.md)** (20ns → 15ns) |\n| [Quill](https://github.com/odygrd/quill) | Lock-free SPSC queue with deferred formatting; only viable approach for hot-path logging without blocking | `[Adopted]` Quill as logging backend | **8ns** median latency; zero blocking |\n| [NanoLog](https://github.com/PlatformLab/NanoLog) | Binary encoding + background thread achieves 7ns; compile-time format validation essential for type safety | `[Synthesized]` `DeferredProcessor\u003cT\u003e` with static type-safe binary serialization | **[84% reduction](docs/compare/DEFERRED_PROCESSOR_BENCHMARK.md)** (75ns → 12ns) |\n| [liburing](https://github.com/axboe/liburing) | `DEFER_TASKRUN` defers completion to userspace, eliminating kernel task wakeups; registered buffers avoid per-op mapping | `[Adopted]` io_uring + DEFER_TASKRUN + registered buffers + multishot | **[7-27% faster](docs/compare/DEFER_TASKRUN_BENCHMARK.md)**; ~30% fewer syscalls |\n| [Highway](https://github.com/google/highway) | Portable SIMD abstraction across AVX2/AVX-512/NEON/SVE; slight overhead vs direct intrinsics | `[Evaluated]` Retained hand-tuned intrinsics for FIX-specific patterns | **13x throughput**; Highway deferred for ARM |\n| [Seastar](https://github.com/scylladb/seastar) | Share-nothing reactor optimal for high-concurrency I/O; high abstraction overhead for single-threaded tick-to-trade paths | `[Influenced]` Extracted core-pinning + lock-free pipelining without framework | **[8% P99 improvement](docs/compare/CPU_AFFINITY_BENCHMARK.md)** (18.8ns → 17.3ns) |\n| [Folly](https://github.com/facebook/folly) | Advanced memory fencing patterns and lock-free primitives; `folly::Function` overhead acceptable for cold path only | `[Influenced]` Native SPSC queue + bit-masking for tag validation | Comparable performance; zero dependency |\n| [Rigtorp](https://github.com/rigtorp/SPSCQueue) | Cache-line padding (`alignas(64)`) eliminates false sharing; simplest correct SPSC implementation | `[Synthesized]` Native `SPSCQueue` with identical techniques | **88M ops/sec**; 11ns median |\n| [xsimd](https://github.com/xtensor-stack/xsimd) | Generic SIMD wrappers useful for math, but FIX parsing requires byte-level shuffle control | `[Evaluated]` Direct Intel intrinsics for SOH/delimiter scanning | **2x faster** than generic wrappers |\n| [Boost.PMR](https://www.boost.org/doc/libs/release/libs/container/doc/html/container/polymorphic_memory_resources.html) | Standard allocators induce non-deterministic jitter; monotonic buffer enables arena allocation per message | `[Adopted]` `std::pmr::monotonic_buffer_resource` | **Zero heap allocation** on hot path |\n\n### What We Built\n\n| Component | Inspired By | Implementation |\n|-----------|-------------|----------------|\n| `TagOffsetMap` | hffix | Compile-time generated O(1) field lookup table |\n| `DeferredProcessor\u003cT\u003e` | NanoLog | SPSC queue + background thread for async processing |\n| `ThreadLocalPool\u003cT\u003e` | NanoLog, Folly | Per-thread object pool, zero lock contention |\n| `SPSCQueue\u003cT\u003e` | Rigtorp, Folly | Cache-line aligned lock-free queue |\n| `simd_scanner` | xsimd (concept) | Hand-tuned AVX2/AVX-512 SOH and delimiter scanning |\n| `IoUringTransport` | liburing | DEFER_TASKRUN + registered buffers + multishot recv |\n| `CpuAffinity` | Seastar | Thread-to-core pinning utility |\n\n### Cumulative Impact\n\n| Metric | Before | After | Improvement |\n|--------|--------|-------|-------------|\n| ExecutionReport Parse | 730 ns | 246 ns | **3.0x faster** |\n| Hot Path Latency | 361 ns | 213 ns | **41% reduction** |\n| SIMD SOH Scan | ~150 ns | 11.8 ns | **~13x faster** |\n| Hash Map Lookup | 20 ns | 15 ns | **31% faster** |\n| P99 Tail Latency | 784 ns | 258 ns | **3.0x lower** |\n\n*Detailed benchmarks: [Optimization Summary](docs/compare/OPTIMIZATION_SUMMARY_BEFORE_AFTER.md)*\n\n### Attribution\n\nNexusFIX is MIT licensed. We gratefully acknowledge these open source projects:\n\n| Dependency | License | Usage |\n|------------|---------|-------|\n| [Abseil](https://github.com/abseil/abseil-cpp) | Apache 2.0 | `flat_hash_map` for session lookups |\n| [Quill](https://github.com/odygrd/quill) | MIT | Async logging infrastructure |\n| [liburing](https://github.com/axboe/liburing) | MIT/LGPL | io_uring C wrapper |\n\n---\n\n## Features\n\n### Core Capabilities\n\n- **Zero-Copy Parsing** - `std::span\u003cconst char\u003e` views into original buffer, no `memcpy`\n- **Message Encoding** - Builder pattern with `constexpr` serializer for constructing FIX messages\n- **SIMD Acceleration** - AVX2/AVX-512 instructions for delimiter scanning\n- **Compile-Time Optimization** - `consteval` field offsets, 22 lookup tables for enum/type conversion, ~300 runtime branches eliminated\n- **O(1) Field Lookup** - Pre-indexed lookup table by FIX tag number (post-parse)\n- **Zero Heap Allocation** - PMR pools and stack allocation on hot path\n- **Session Management** - Full session lifecycle: Logon, Logout, Heartbeat, sequence number tracking, reconnect logic\n- **Type-Safe API** - Strong types for Price, Quantity, Side, OrdType\n\n### Modern C++23\n\n- `std::expected` for error handling (no exceptions on hot path)\n- `std::span` for zero-copy data views\n- Concepts for compile-time interface validation\n- `consteval` for compile-time computation\n- `[[likely]]` / `[[unlikely]]` branch hints\n\n### Supported FIX Versions\n\n| Version | Status | Notes |\n|---------|--------|-------|\n| FIX 4.4 | Full Support | Most common in production |\n| FIX 5.0 + FIXT 1.1 | Full Support | Only 2% overhead vs 4.4 |\n\n### Supported Message Types\n\n| MsgType | Name | Category |\n|---------|------|----------|\n| A | Logon | Session |\n| 5 | Logout | Session |\n| 0 | Heartbeat | Session |\n| D | NewOrderSingle | Order Entry |\n| F | OrderCancelRequest | Order Entry |\n| 8 | ExecutionReport | Order Entry |\n| V | MarketDataRequest | Market Data |\n| W | MarketDataSnapshotFullRefresh | Market Data |\n| X | MarketDataIncrementalRefresh | Market Data |\n\n### Optimization Guide\n\nHow we achieved sub-300ns latency with Modern C++23:\n\n- [Optimization Diary](docs/optimization_diary.md) - Step-by-step journey from 730ns to 246ns\n- [Modern C++ Quant Techniques](docs/modernc_quant.md) - Cache-line alignment, SIMD, PMR strategies, branch hints\n\n---\n\n## Quick Start\n\n### Installation\n\n```bash\ngit clone https://github.com/Lattice9AI/NexusFIX.git\ncd NexusFIX\n./start.sh build\n```\n\n### Requirements\n\n- **C++23 compiler**: GCC 13+ or Clang 17+\n- **CMake**: 3.20+\n- **OS**: Linux (io_uring optional), macOS, Windows\n\n### Basic Usage\n\n```cpp\n#include \u003cnexusfix/nexusfix.hpp\u003e\n\nusing namespace nfx;\nusing namespace nfx::fix44;\n\n// Connect to broker\nTcpTransport transport;\ntransport.connect(\"fix.broker.com\", 9876);\n\n// Configure session\nSessionConfig config{\n    .sender_comp_id = \"MY_CLIENT\",\n    .target_comp_id = \"BROKER\",\n    .heartbeat_interval = 30\n};\nSessionManager session{transport, config};\nsession.initiate_logon();\n\n// Send order (zero allocation)\nMessageAssembler asm_;\nNewOrderSingle::Builder order;\nauto msg = order\n    .cl_ord_id(\"ORD001\")\n    .symbol(\"AAPL\")\n    .side(Side::Buy)\n    .order_qty(Qty::from_int(100))\n    .ord_type(OrdType::Limit)\n    .price(FixedPrice::from_double(150.00))\n    .build(asm_);\ntransport.send(msg);\n```\n\n### Parse Incoming Messages\n\n```cpp\n// Zero-copy parsing\nFixParser parser;\nauto result = parser.parse(raw_buffer);\n\nif (result) {\n    auto\u0026 msg = *result;\n    auto order_id = msg.get_string(Tag::OrderID);    // O(1) lookup\n    auto exec_type = msg.get_char(Tag::ExecType);    // No allocation\n    auto fill_qty = msg.get_qty(Tag::LastQty);       // Type-safe\n}\n```\n\n---\n\n## Build Options\n\n| CMake Option | Default | Description |\n|--------------|---------|-------------|\n| `NFX_ENABLE_SIMD` | ON | AVX2/AVX-512 SIMD acceleration |\n| `NFX_ENABLE_IO_URING` | OFF | Linux io_uring transport |\n| `NFX_BUILD_BENCHMARKS` | ON | Build benchmark suite |\n| `NFX_BUILD_TESTS` | ON | Build unit tests |\n| `NFX_BUILD_EXAMPLES` | ON | Build examples |\n\n```bash\n# Build with all optimizations\ncmake -B build -DCMAKE_BUILD_TYPE=Release -DNFX_ENABLE_SIMD=ON\ncmake --build build -j\n\n# Run benchmarks\n./start.sh bench 100000\n\n# Compare with QuickFIX\n./start.sh compare 100000\n```\n\n---\n\n## Benchmarking\n\nVerify performance claims by running benchmarks yourself.\n\n### Quick Start\n\n```bash\n# Run parser and session benchmarks\n./start.sh bench 100000\n\n# Example output:\n# [BENCHMARK] ExecutionReport Parse\n#   Iterations: 100000\n#   Mean: 246 ns\n#   P50:  245 ns\n#   P99:  258 ns\n```\n\n### QuickFIX Comparison\n\nCompare NexusFIX against QuickFIX (requires QuickFIX installed):\n\n```bash\n# Install QuickFIX first\n# Ubuntu: sudo apt install libquickfix-dev\n# Or build from source: https://github.com/quickfix/quickfix\n\n# Run comparison\n./start.sh compare 100000\n```\n\n### Full Reproduction Guide\n\nFor detailed instructions on reproducing benchmark results, including:\n- Environment setup (CPU governor, pinning, priority)\n- Build configuration options\n- Interpreting results\n- Troubleshooting\n\nSee [BENCHMARK_REPRODUCTION.md](BENCHMARK_REPRODUCTION.md)\n\n---\n\n## Documentation\n\n- [API Reference](docs/API_REFERENCE.md) - Complete API documentation\n- [Implementation Guide](docs/design/IMPLEMENTATION_GUIDE.md) - Architecture overview\n- [Benchmark Report](docs/compare/BENCHMARK_COMPARISON_REPORT.md) - Detailed performance analysis\n- [Modern C++ Techniques](docs/modernc_quant.md) - Optimization techniques used\n\n---\n\n## Project Structure\n\n```\nnexusfix/\n├── include/nexusfix/\n│   ├── parser/           # Zero-copy FIX parser (SIMD)\n│   ├── session/          # Session state machine\n│   ├── transport/        # TCP / io_uring / Winsock transport\n│   ├── platform/         # Cross-platform abstraction\n│   ├── types/            # Strong types (Price, Qty, Side)\n│   ├── memory/           # PMR buffer pools\n│   ├── store/            # Message store (PMR-optimized)\n│   ├── serializer/       # Message serialization\n│   ├── util/             # Utilities (diagnostics, formatting)\n│   ├── messages/fix44/   # FIX 4.4 message builders\n│   └── interfaces/       # Concepts and interfaces\n├── benchmarks/           # Performance benchmarks\n├── tests/                # Unit tests\n├── examples/             # Example programs\n└── docs/                 # Documentation\n```\n\n---\n\n## FAQ\n\n### How does NexusFIX achieve zero-copy parsing?\n\nNexusFIX uses `std::span\u003cconst char\u003e` to create views into the original network buffer. Field values are never copied - the parser returns spans pointing to the exact byte range in the source buffer. This eliminates all `memcpy` and heap allocation overhead.\n\n### Is NexusFIX compatible with QuickFIX?\n\nNexusFIX implements the same FIX 4.4/5.0 protocol standards but with a different API optimized for performance. It is wire-compatible with any FIX counterparty, including systems using QuickFIX.\n\n### What latency can I expect in production?\n\nIn our benchmarks: **~250 nanoseconds** for ExecutionReport parsing. Actual production latency depends on network, kernel configuration, and hardware. NexusFIX is designed to minimize the application-layer overhead.\n\n### Does NexusFIX support FIX Repeating Groups?\n\nYes. Repeating groups are parsed with the same zero-copy approach. Group iteration is O(1) per entry.\n\n---\n\n## Use Cases\n\nNexusFIX is designed for:\n\n- **High-Frequency Trading (HFT)** - Sub-microsecond message processing\n- **Algorithmic Trading Systems** - Low-latency order routing\n- **Market Making** - High-throughput quote updates\n- **Smart Order Routing (SOR)** - Multi-venue connectivity\n- **Trading Infrastructure** - FIX gateways and bridges\n\n---\n\n## Contact\n\nFor questions or collaboration: nonagonal.portal@gmail.com\n\n---\n\n## Development\n\nBuilt with **Modern C++23**. Optimized via hardware-aware high-performance patterns including cache-line alignment, SIMD vectorization, and zero-copy memory design. Verified through rigorous benchmarking and AI-assisted static analysis.\n\nFor technical deep-dives on our optimization journey, see [Optimization Diary](docs/optimization_diary.md).\n\n---\n\n## Contributing\n\nThis project is maintained by **Lattice9AI**.\n\n- **Issues \u0026 Discussions**: Welcome for bug reports, performance questions, and feature discussions\n- **Pull Requests**: Bug fixes and performance optimizations welcome (see [CONTRIBUTING.md](CONTRIBUTING.md))\n- Feature PRs require prior discussion in Issues\n- Performance PRs must include benchmark data (before/after)\n\nAll contributions must follow:\n- C++23 standards\n- Zero allocation on hot paths\n- Include benchmarks for performance changes\n\n---\n\n## License\n\nMIT License - See [LICENSE](LICENSE) file.\n\n---\n\n\u003cp align=\"center\"\u003e\n  \u003csub\u003eBuilt with Modern C++23 for ultra-low latency quantitative trading\u003c/sub\u003e\n\u003c/p\u003e\n","funding_links":[],"categories":["Trading \u0026 Backtesting","编程"],"sub_categories":["C++"],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FLattice9AI%2FNexusFix","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FLattice9AI%2FNexusFix","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FLattice9AI%2FNexusFix/lists"}