Datara Logo Datara
Features Performance Language Reference Tutorial All Docs Install GitHub
Datara v1.4.5 Production Release

The Post-OOP Systems Language
for Predictable Microsecond Speed

Zero GC pauses, affine memory safety, mathematical Evidence Gate optimizer, and dual Cranelift + LLVM AOT native code generation.

Download Datara-Setup.exe (64-bit) View Source on GitHub
Quick Install Terminal
# Windows PowerShell (Dynamic Latest Release)
>irm https://raw.githubusercontent.com/datara-lang/datara/main/install.ps1 | iex
# Linux & macOS (Bash)
$curl -fsSL https://raw.githubusercontent.com/datara-lang/datara/main/install.sh | bash

Engineered for Mechanical Sympathy

Designed from first principles to extract maximum performance from modern x86_64 and ARM64 CPUs.

Dual-Engine Codegen

Sub-50ms cold compilation with Cranelift JIT for rapid local development. Switch to --llvm for LLVM -O3, LTO, and adaptive SIMD vectorization.

Evidence Gate Formal Optimizer

Transformations are backed by structural mathematical proof. Countable loops are solved in O(1) time, while structs live inside CPU registers via SROA.

Post-OOP Data-Oriented Design

Clean separation of data records (entity, packet) from execution logic (behavior). Direct dispatch compiles to raw CALLs without vtables.

Zero GC Pauses

Deterministic affine ownership semantics and zero-copy borrowing (view). Never suffer garbage collection latency spikes in high-load services.

Universal CPU Portability

Generates code adhering to guaranteed baseline instruction sets (SSE2/NEON), ensuring binaries run seamlessly across 100% of computers without crashing.

All-in-One Developer Toolchain

forgen contains everything: compiler, JIT REPL, test runner, code formatter, linter, LSP server, and C99 header exporter.

Verified Performance & Optimization Matrix

Real measured execution metrics across 10 core workloads executing through Datara's Evidence Gate optimizer.

Workload Category Dataset / Trips Evidence Gate Pass Baseline Cranelift JIT LLVM AOT Gain / Complexity Allocations
Integer Arithmetic Loop 10,000,000 trips Closed-Form Arithmetic Reduction 14.80 ms 0.00 ms 0.00 ms Instant O(1) Fold 0 bytes
Float Polynomial Compute 1,000,000 points Horner Induction Reassociation 8.40 ms 0.00 ms 0.00 ms Instant O(1) Fold 0 bytes
Struct 2D/3D Vector Math 1,000,000 structs Mutable SROA (Scalar Replacement) 19.20 ms 0.00 ms 0.00 ms Register Resident 0 bytes
Post-OOP Method Dispatch 1,000,000 calls Monomorphic Inlining & Direct Call 12.60 ms 5.65 ms 2.10 ms Direct Call (No Vtable) 0 bytes
Generic Box Operations 1,000,000 items Zero-Cost Box Monomorphization 16.50 ms 0.00 ms 0.00 ms Zero Box Overhead 0 bytes
Pipeline Dataflow (|>) 1,000,000 items Polyhedral Stream Operator Fusion 22.00 ms 0.00 ms 0.00 ms Single-Pass Fusion 0 bytes
Array Vectorized Compute 1,000,000 items Adaptive SIMD (8x AVX2 / 4x SSE2) 4.10 ms 0.08 ms 0.04 ms 102.5x Vectorized 0 bytes
String Wire-Blit Fusion 250,000 strings Polyhedral Splice & Exact Sizing 38.00 ms 0.00 ms 0.00 ms Instant Wire-Blit 0 bytes