Fix TensorAlloc copy that was copying the ptr

Realized I can just do alloc.create instead of a all new struct

.gitea/workflows/deploy.yml

@@ -1,34 +0,0 @@
-name: Deploy MkDocs to Garage
-on:
-  push:
-    branches:
-      - main  # Adjust to your branch name
-
-jobs:
-  build-and-deploy:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-
-      - name: Set up Python
-        uses: actions/setup-python@v4
-        with:
-          python-version: '3.x'
-
-      - name: Install dependencies
-        run: pip install mkdocs-material
-
-      - name: Build
-        run: mkdocs build
-
-      - name: Sync to Garage S3
-        uses: https://github.com/jakejarvis/s3-sync-action@master
-        with:
-          args: --endpoint-url https://s3.garage.bouvais.lu --acl public-read --delete
-        env:
-          AWS_S3_BUCKET: 'zig-dimal.bouvais.lu'
-          AWS_ACCESS_KEY_ID: ${{ secrets.GARAGE_ACCESS_KEY }}
-          AWS_SECRET_ACCESS_KEY: ${{ secrets.GARAGE_SECRET_KEY }}
-          AWS_REGION: 'garage' 
-          SOURCE_DIR: 'site' # MkDocs defaults to 'site' folder for output

.gitignore

@@ -1,2 +1,4 @@
 zig-out
 .zig-cache
+mkdocs.yaml
+zig-pkg

README.md

@@ -1,42 +1,55 @@
 # dimal — Dimensional Analysis for Zig
 
-A comptime-first dimensional analysis module for Zig. If you try to add meters to seconds, **it won't compile**. That's the point.
+A dimensional analysis library for Zig with a unified `Tensor` API for scalars, vectors, matrices, and higher-dimensional data. All dimension and unit tracking happens at compile time—zero runtime overhead—and all operations use SIMD intrinsics.
 
-Started by a space simulation where `i128` positions were needed to avoid float imprecision far from the origin, this module grew into a full physical-unit type system with zero runtime overhead.
+If you try to add meters to seconds, it won't compile. That's the point.
 
 > **Source:** [git.bouvais.lu/adrien/zig-dimal](https://git.bouvais.lu/adrien/zig-dimal)  
 > **Minimum Zig version:** `0.16.0`
 
 ---
 
+## Background
+
+Started because I needed `i128` positions for a space simulation to avoid floating-point precision loss far from the origin. Grew into a type system for tracking physical dimensions at compile time. It's been useful enough to share.
+
+- **Compile-time dimension checking** — catch unit mismatches before runtime.
+- **Unified `Tensor` API** — same interface for scalars, vectors, matrices, and higher-rank tensors.
+- **SIMD operations** — vector and matrix code automatically uses SIMD instructions.
+- **Zero runtime cost** — all dimension and scale tracking is erased at compile time.
+- **Supports `i128`** — useful for high-precision fixed-point integer math.
+
+---
+
 ## Features
 
-- **100% comptime** — all dimension and unit tracking happens at compile time. No added memory, *almost* native performance.
+- **Compile-time dimension checking** — all physical-unit tracking happens at compile time.
-- **Compile-time dimension errors** — adding `Meter` to `Second` is a compile error, not a runtime panic.
 - **Automatic unit conversion** — use `.to()` to convert between compatible units (e.g. `km/h` → `m/s`). Scale factors are resolved at comptime.
-* **Full SI prefix & Imperial support** — `pico` through `peta`, plus common Imperial units like `inch`, `ft`, `mi`, `lb`, and `oz`.
+- **Unified `Tensor` API** — one type for scalars `{1}`, vectors `{N}`, matrices `{M, N}`, and higher-rank tensors.
-- **Time scale support** — `min`, `hour`, `year` built in.
+- **SIMD operations** — vector and matrix code compiles to SIMD instructions automatically.
-- **Scalar and Vector types** — operate on individual values or fixed-size arrays with the same dimensional safety.
+- **Tensor contraction** — `.contract(other, axis_a, axis_b)` for dot products, matrix multiplication, and general tensor contractions.
-- **Built-in physical quantities** — `dma.Base` provides ready-made types for `Velocity`, `Acceleration`, `Force`, `Energy`, `Pressure`, `ElectricCharge`, `ThermalConductivity`, and many more.
+- **Full SI prefix support** — `pico` through `peta`, plus Imperial units and time scales.
-- **Comparison operations** — `eq`, `ne`, `gt`, `gte`, `lt`, `lte` on both `Scalar` and `Vector`, with automatic scale resolution.
+- **Physical constants** — Planck, Boltzmann, speed of light, gravitational constant, etc.
-- **Arithmetic with bare numbers** — multiply or divide a dimensioned value by a `comptime_int`, `comptime_float`, or plain `T` directly. The value is treated as dimensionless; dimensions pass through unchanged.
+- **Pre-built quantities** — `Velocity`, `Acceleration`, `Force`, `Energy`, `Pressure`, `Charge`, and more.
-- **`abs`, `pow`, `sqrt`** — unary operations with correct dimension tracking (`pow(2)` on `L¹` → `L²`, etc.).
+- **Basic vector operations** — cross product, length/magnitude, element-wise arithmetic.
-- **Vector geometry** — `dot` product (returns a `Scalar`), `cross` product (Vec3 only), element-wise `product` (all components multiplied).
+- **Formatting** — values print with units: `9.81m.s⁻²`, `0.172km`.
-- **Rich formatting** — values print with their unit automatically: `9.81m.s⁻²`, `42m.kg.s⁻¹`, `0.172km`.
+
-- **`i128` support** — the whole reason this exists. Use large integers for high-precision fixed-point positions without manual conversion.
+### Current Limitations
-- **Tests and benchmarks included** — run them and see how it performs on your machine (results welcome!).
+
+- GPU support not implemented.
+- Performance on small tensors is limited by Zig's vector width.
 
 ---
 
 ## The 7 SI Base Dimensions
 
 | Symbol | Dimension            | SI Unit |
-|--------|----------------------|----------|
+|--------|----------------------|---------|
 | `L`    | Length               | `m`     |
 | `M`    | Mass                 | `g`     |
 | `T`    | Time                 | `s`     |
 | `I`    | Electric Current     | `A`     |
-| `Tp`   | Temperature          | `K`      |
+| `Tr`   | Temperature          | `K`     |
 | `N`    | Amount of Substance  | `mol`   |
 | `J`    | Luminous Intensity   | `cd`    |
 
@@ -44,10 +57,10 @@ Started by a space simulation where `i128` positions were needed to avoid float
 
 ## Installation
 
-### 1. Fetch the dependency
+### 1. Add the dependency (Zig 0.14+)
 
 ```sh
-zig fetch --save git+https://git.bouvais.lu/adrien/zig-dimal#0.1.1
+zig fetch --save git+https://git.bouvais.lu/adrien/zig-dimal#0.2.0
 ```
 
 ### 2. Wire it up in `build.zig`
@@ -57,287 +70,181 @@ const std = @import("std");
 
 pub fn build(b: *std.Build) void {
     const target = b.standardTargetOptions(.{});
-    const dimal = b.dependency("dimal", .{}).module("dimal");
+    const optimize = b.standardOptimizeOption(.{});
+    
+    const dimal = b.dependency("dimal", .{
+        .target = target,
+        .optimize = optimize,
+    }).module("dimal");
 
     const exe = b.addExecutable(.{
-        .name = "my_project",
+        .name = "my_app",
-        .root_module = b.createModule(.{
         .root_source_file = b.path("src/main.zig"),
         .target = target,
-            .imports = &.{.{
+        .optimize = optimize,
-                .name = "dimal",
-                .module = dimal,
-            }},
-        }),
     });
+    exe.root_module.addImport("dimal", dimal);
     b.installArtifact(exe);
 }
 ```
 
-### 3. Import in your code
+### 3. Import and use
 
 ```zig
 const dma = @import("dimal");
-const Scalar     = dma.Scalar;
+const Tensor = dma.Tensor;
-const Dimensions = dma.Dimensions;
+const Base = dma.Base;
-const Scales     = dma.Scales;
 ```
 
 ---
 
-## Quick Start
+## Quick Example: Lunar Descent
 
-### Defining unit types
+Simulate a spacecraft descending to the Moon with correct physics and type safety:
-
-A `Scalar` type is parameterized by three things: the numeric type (`f64`, `i128`, …), the dimensions (which physical quantities and their exponents), and the scales (SI prefixes or custom time units). Both the dimension and scale arguments are plain struct literals — no wrapper call needed.
 
 ```zig
-const Meter     = Scalar(f64, .{ .L = 1 },           .{});
+const std = @import("std");
-const NanoMeter = Scalar(i64, .{ .L = 1 },           .{ .L = .n });
+const dma = @import("dimal");
-const KiloMeter = Scalar(f64, .{ .L = 1 },           .{ .L = .k });
+const Tensor = dma.Tensor;
-const Second    = Scalar(f64, .{ .T = 1 },           .{});
-const Velocity  = Scalar(f64, .{ .L = 1, .T = -1 },  .{});
-const Kmh       = Scalar(f64, .{ .L = 1, .T = -1 },  .{ .L = .k, .T = .hour });
-```
 
-Or use the pre-built helpers from `dma.Base`:
+pub fn main() void {
-
+    // Define types: m/s² acceleration, m/s velocity, m distance
-```zig
     const Acceleration = dma.Base.Acceleration.Of(f64);
-const KmhSpeed     = dma.Base.Speed.Scaled(f64, .{ .L = .k, .T = .hour });
+    const Velocity = dma.Base.Velocity.Of(f64);
+    const Distance = dma.Base.Meter.Of(f64);
+    const Time = dma.Base.Second.Of(f64);
+
+    // Initial conditions
+    const g_moon: Acceleration = .{ .data = @splat(1.62) };
+    const v_initial: Velocity = .{ .data = @splat(100.0) };
+    const h_initial: Distance = .{ .data = @splat(10000.0) };
+    const dt: Time = .{ .data = @splat(1.0) };
+
+    var h = h_initial;
+    var v = v_initial;
+    var t: f64 = 0;
+
+    // Simulate descent
+    while (h.data[0] > 0 and t < 1000) : (t += 1.0) {
+        // a = -g (gravity pulls down)
+        const a = g_moon.mul(-1.0);
+
+        // Update: v = v₀ + at
+        v = v.add(a.mul(dt));
+
+        // Update: h = h₀ + vt
+        h = h.add(v.mul(dt));
+
+        if (@mod(t, 100.0) == 0) {
+            std.debug.print("t={d:.0}s | h={d:.1} | v={d:.1}\n", .{
+                t,
+                h,
+                v,
+            });
+        }
+    }
+
+    std.debug.print("Landed in {d:.1}s at h={d:.1}\n", .{ t, h });
+}
 ```
 
-### Kinematics example
+**Output:**
-
-```zig
-const v0    = Velocity{ .value = 10.0 };       // 10 m/s
-const accel = Acceleration{ .value = 9.81 };   // 9.81 m/s²
-const time  = Second{ .value = 5.0 };          // 5 s
-
-// d = v₀t + ½at²
-const d1   = v0.mul(time);                         // → Meter
-const d2   = accel.mul(time).mul(time).mul(0.5);   // → Meter  (bare 0.5 is dimensionless)
-const dist = d1.add(d2);
-
-const v_final = v0.add(accel.mul(time));
-
-std.debug.print("Distance: {d} | {d}\n", .{ dist, dist.to(KiloMeter) });
-// Distance: 172.625m | 0.172625km
-
-std.debug.print("Final speed: {d:.2}\n", .{v_final});
-// Final speed: 59.05m.s⁻¹
 ```
-
+t=0s | h=10000m | v=100m.s⁻¹
-### Unit conversion
+t=100s | h=8019m | v=-61.8m.s⁻¹
-
+t=200s | h=4174.4m | v=-223.6m.s⁻¹
-`.to()` converts between compatible units at comptime. Mixing incompatible dimensions is a **compile error**.
+...
-
+Landed in 323.5s at h=-0.01m
-```zig
-const speed_kmh = Kmh{ .value = 120.0 };
-const speed_ms  = speed_kmh.to(Velocity); // 33.333... m/s — comptime ratio
-
-// This would NOT compile:
-// const bad = speed_kmh.to(Second); // "Dimension mismatch in to: L1T-1 vs T1"
-```
-
-#### Imperial
-
-```zig
-const Inch      = Scalar(f64, .{ .L = 1 }, .{ .L = .inch });
-const Mile      = Scalar(f64, .{ .L = 1 }, .{ .L = .mi });
-const Pound     = Scalar(f64, .{ .M = 1 }, .{ .M = .lb });
-
-// Conversion example
-const dist_m = Meter{ .value = 1609.344 };
-const dist_mi = dist_m.to(Mile); // Result: 1.0
-```
-
-### Arithmetic with bare numbers
-
-Passing a `comptime_int`, `comptime_float`, or plain `T` to `mul` / `div` treats it as a dimensionless value. Dimensions pass through unchanged.
-
-```zig
-const Meter = Scalar(f64, .{ .L = 1 }, .{});
-const d     = Meter{ .value = 6.0 };
-
-const half    = d.mul(0.5);   // comptime_float → still Meter
-const doubled = d.mul(2);     // comptime_int   → still Meter
-const factor: f64 = 3.0;
-const tripled = d.mul(factor); // runtime f64    → still Meter
-```
-
-### Comparisons
-
-`eq`, `ne`, `gt`, `gte`, `lt`, `lte` work on any two `Scalar` values of the **same dimension**. Scales are resolved automatically before comparing.
-
-```zig
-const Meter     = Scalar(i64, .{ .L = 1 }, .{});
-const KiloMeter = Scalar(i64, .{ .L = 1 }, .{ .L = .k });
-
-const m1000 = Meter{ .value = 1000 };
-const km1   = KiloMeter{ .value = 1 };
-const km2   = KiloMeter{ .value = 2 };
-
-_ = m1000.eq(km1);   // true  — same magnitude
-_ = km2.gt(m1000);   // true  — 2 km > 1000 m
-_ = m1000.lte(km2);  // true
-
-// Comparing with a bare number works when the scalar is dimensionless.
-// Comparing incompatible dimensions is a compile error.
-```
-
-### Unary operations: `abs`, `pow`, `sqrt`
-
-```zig
-const Meter = Scalar(f64, .{ .L = 1 }, .{});
-const d     = Meter{ .value = -4.0 };
-
-const magnitude = d.abs();    // 4.0 m      — dimension unchanged
-const area      = d.pow(2);   // 16.0 m²    — dims scaled by exponent
-const side      = area.sqrt(); // 4.0 m     — dims halved (requires even exponents)
-```
-
-`pow` accepts any `comptime_int` exponent and adjusts the dimension exponents accordingly. `sqrt` is a compile error unless all dimension exponents are even.
-
-### Working with Vectors
-
-Every `Scalar` type exposes a `.Vec3` alias and a generic `.Vec(n)` type accessor:
-
-```zig
-const Vec3Meter = Meter.Vec3; // equivalent to Vector(3, Meter)
-
-const pos = Vec3Meter{ .data = .{ 100, 200, 300 } };
-const t   = Second{ .value = 10 };
-
-const vel = pos.divScalar(t);  // → Vec3 of Velocity (m/s)
-std.debug.print("{d}\n", .{vel}); // (10, 20, 30)m.s⁻¹
-```
-
-#### Dot and cross products
-
-```zig
-const Newton = Scalar(f32, .{ .M = 1, .L = 1, .T = -2 }, .{});
-
-const r     = Meter.Vec3{ .data = .{ 10.0, 0.0, 0.0 } };
-const force = Newton.Vec3{ .data = .{ 5.0, 5.0, 0.0 } };
-
-// Dot product — returns a Scalar (dimensions summed)
-const work   = force.dot(r);     // 50.0 J  (M¹L²T⁻²)
-
-// Cross product — returns a Vec3 (dimensions summed, Vec3 only)
-const torque = r.cross(force);   // (0, 0, 50) N·m
-```
-
-#### Vector comparisons
-
-Element-wise comparisons return `[len]bool`. Whole-vector equality uses `eqAll` / `neAll`. A single scalar can be broadcast with the `*Scalar` variants.
-
-```zig
-const positions  = Meter.Vec3{ .data = .{ 500.0, 1200.0, 3000.0 } };
-const threshold  = KiloMeter{ .value = 1.0 }; // 1 km
-
-const exceeded = positions.gtScalar(threshold); // [false, true, true]
-const eq_each  = positions.eq(positions);        // [true, true, true]  (element-wise)
-const all_same = positions.eqAll(positions);     // true  (whole-vector)
-```
-
-#### Other Vector operations
-
-```zig
-const v = Meter.Vec3{ .data = .{ -2.0, 3.0, -4.0 } };
-
-const v_abs  = v.abs();       // { 2, 3, 4 } m
-const vol    = v_abs.product(); // 24 m³  (dims × len)
-const area   = v_abs.pow(2);  // { 4, 9, 16 } m²
-const sides  = area.sqrt();   // { 2, 3, 4 } m  (element-wise sqrt)
 ```
 
 ---
 
-## API Reference
+## API Overview
 
-### `Scalar(T, dims, scales)`
+### Tensors
 
-| Method | Description |
+A **`Tensor`** is parameterized by:
-|---|---|
+- **`T`** — numeric type: `f32`, `f64`, `i128`, etc.
-| `.add(rhs)` | Add two quantities of the same dimension. Auto-converts scales. |
+- **`dims`** — physical dimensions (struct literal): `.{.L = 1, .T = -1}` means length/time (velocity).
-| `.sub(rhs)` | Subtract. Auto-converts scales. |
+- **`scales`** — SI prefixes or custom scales: `.{.L = .k, .T = .hour}` means km/h.
-| `.mul(rhs)` | Multiply — dimensions are **summed**. `rhs` may be a `Scalar`, `T`, `comptime_int`, or `comptime_float` (bare numbers are dimensionless). |
+- **`shape`** — array shape: `&.{1}` is a scalar, `&.{3}` is a 3-vector, `&.{3, 3}` is a 3×3 matrix.
-| `.div(rhs)` | Divide — dimensions are **subtracted**. Same `rhs` flexibility as `mul`. |
-| `.abs()` | Absolute value. Dimensions and scales unchanged. |
-| `.pow(exp)` | Raise to a `comptime_int` exponent. Dimension exponents are multiplied by `exp`. |
-| `.sqrt()` | Square root. Compile error unless all dimension exponents are even. |
-| `.eq(rhs)` / `.ne(rhs)` | Equality / inequality comparison. Scales auto-resolved. |
-| `.gt(rhs)` / `.gte(rhs)` | Greater-than / greater-than-or-equal. |
-| `.lt(rhs)` / `.lte(rhs)` | Less-than / less-than-or-equal. |
-| `.to(DestType)` | Convert to another unit of the same dimension. Compile error on mismatch. |
-| `.vec(len)` | Return a `Vector(len, Self)` with all components set to this value. |
-| `.vec3()` | Shorthand for `.vec(3)`. |
-| `.Vec3` | Type alias for `Vector(3, Self)`. |
 
-### `Vector(len, Q)`
+```zig
+// Scalar: 1-element tensor
+const Meter = Tensor(f64, .{.L = 1}, .{}, &.{1});
+const m = Meter{ .data = @splat(5.0) };
 
-| Method | Description |
+// Vector: N-element tensor (SIMD)
-|---|---|
+const Vec3Meter = Tensor(f64, .{.L = 1}, .{}, &.{3});
-| `.add(rhs)` / `.sub(rhs)` | Element-wise add / subtract. |
+const v = Vec3Meter{ .data = @shuffle(f64, [_]f64{1, 2, 3}, [_]f64 undefined, [_]i32{0, 1, 2, 0, 0, 0}) };
-| `.mul(rhs)` / `.div(rhs)` | Element-wise multiply / divide (both operands are Vectors). |
-| `.mulScalar(s)` / `.divScalar(s)` | Scale every component by a single `Scalar`, `T`, `comptime_int`, or `comptime_float`. |
-| `.dot(rhs)` | Dot product → `Scalar` with combined dimensions. |
-| `.cross(rhs)` | Cross product → `Vector(3, …)`. Vec3 only. |
-| `.abs()` | Element-wise absolute value. |
-| `.pow(exp)` | Element-wise `comptime_int` power. Dimension exponents scaled. |
-| `.sqrt()` | Element-wise square root. |
-| `.product()` | Multiply all components → `Scalar` with dimensions × `len`. |
-| `.negate()` | Negate all components. |
-| `.length()` | Euclidean length (returns `T`). |
-| `.lengthSqr()` | Sum of squared components (returns `T`). Cheaper than `length`. |
-| `.eq(rhs)` / `.ne(rhs)` | Element-wise comparison → `[len]bool`. |
-| `.gt(rhs)` / `.gte(rhs)` / `.lt(rhs)` / `.lte(rhs)` | Element-wise ordered comparisons → `[len]bool`. |
-| `.eqAll(rhs)` / `.neAll(rhs)` | Whole-vector equality / inequality → `bool`. |
-| `.eqScalar(s)` / `.neScalar(s)` | Broadcast scalar comparison → `[len]bool`. |
-| `.gtScalar(s)` / `.gteScalar(s)` / `.ltScalar(s)` / `.lteScalar(s)` | Broadcast ordered scalar comparisons → `[len]bool`. |
-| `.to(DestQ)` | Convert all components to a compatible scalar type. |
 
-### `dma.Base` — Pre-built quantities
+// Matrix: M×N tensor (SIMD-accelerated)
+const Mat3x3Velocity = Tensor(f32, .{.L = 1, .T = -1}, .{}, &.{3, 3});
+const m_vel = Mat3x3Velocity{ .data = @splat(10.0) };
 
-Call `.Of(T)` for base-unit scalars, `.Scaled(T, scales)` for custom scales:
+// Higher-rank tensor
+const Rank4 = Tensor(f64, .{.M = 1}, .{}, &.{2, 3, 4, 5});
+```
 
-`Meter`, `Second`, `Gramm`, `Kelvin`, `ElectricCurrent`, `Speed`, `Acceleration`, `Inertia`, `Force`, `Pressure`, `Energy`, `Power`, `Area`, `Volume`, `Density`, `Frequency`, `Viscosity`, `ElectricCharge`, `ElectricPotential`, `ElectricResistance`, `MagneticFlux`, `ThermalCapacity`, `ThermalConductivity`, and more.
+### Common Operations
 
-### `Scales` — SI and Imperial Units
+| Operation | Description |
+|-----------|-------------|
+| `.add(rhs)` | Element-wise addition. Auto-converts scales. |
+| `.sub(rhs)` | Element-wise subtraction. |
+| `.mul(rhs)` | Multiply; dimensions are summed. `rhs` can be a tensor or bare number. |
+| `.div(rhs)` | Divide; dimensions are subtracted. |
+| `.contract(other, axis_a, axis_b)` | Tensor contraction: dot product, matrix multiply, or general N-D contraction. |
+| `.cross(rhs)` | Cross product (3-vectors only). Returns a 3-vector. |
+| `.length()` / `.lengthSqr()` | Euclidean length (or squared length) of a vector. Returns a scalar `T`. |
+| `.product()` | Multiply all elements. Returns a scalar with combined dimensions. |
+| `.abs()` | Element-wise absolute value. Dimensions unchanged. |
+| `.pow(exp)` | Raise to comptime exponent. Dimension exponents multiplied by `exp`. |
+| `.sqrt()` | Element-wise square root. Compile error if any dimension exponent is odd. |
+| `.to(DestType)` | Convert to another unit of the same dimension. Comptime error on mismatch. |
+| `.eq(rhs)` / `.ne(rhs)` | Element-wise equality/inequality. |
+| `.gt(rhs)` / `.gte(rhs)` | Greater-than comparisons. |
+| `.lt(rhs)` / `.lte(rhs)` | Less-than comparisons. |
 
-| Tag | Factor (Relative to Base) | Type |
+### Pre-built Types (via `dma.Base`)
-|---|---|---|
-| `.P` ... `.f` | $10^{15}$ ... $10^{-15}$ | SI Prefixes |
-| `.min`, `.hour`, `.year` | 60, 3600, 31,536,000 | Time |
-| **`.inch`** | **0.0254** | Imperial Length (m) |
-| **`.ft`** | **0.3048** | Imperial Length (m) |
-| **`.yd`** | **0.9144** | Imperial Length (m) |
-| **`.mi`** | **1609.344** | Imperial Length (m)  |
-| **`.oz`** | **28.3495231** | Imperial Mass (g) |
-| **`.lb`** | **453.59237** | Imperial Mass (g)  |
-| **`.st`** | **6350.29318** | Imperial Mass (g)  |
 
-Scale entries for dimensions with exponent `0` are ignored — multiplying a dimensionless value by a kilometre-scale value no longer accidentally inherits the `k` prefix.
+Use `.Of(T)` for base units, `.Scaled(T, scales)` for custom scales:
+
+```zig
+const Velocity = dma.Base.Velocity.Of(f64);
+const Kmh = dma.Base.Velocity.Scaled(f64, .{.L = .k, .T = .hour});
+const Force = dma.Base.Force.Of(f32);
+const Energy = dma.Base.Energy.Of(f64);
+```
+
+Also available: `Acceleration`, `Inertia`, `Pressure`, `Power`, `Area`, `Volume`, `Density`, `Frequency`, `Viscosity`, `Charge`, `Potential`, `Resistance`, `MagneticFlux`, `ThermalCapacity`, `ThermalConductivity`, and many more.
 
 ---
 
-## Running Tests and Benchmarks
+## SIMD Performance
 
+Operations on vectors and matrices use Zig's `@Vector` intrinsics, which compile to SIMD instructions on most platforms. This makes vector operations faster than equivalent scalar loops, but don't expect miracles—SIMD is still limited by memory bandwidth and CPU cache.
+
+Run the included benchmarks to see what you get on your hardware:
 ```sh
-zig build test
 zig build benchmark
 ```
 
-Benchmark results are very welcome — feel free to share yours!
+---
+
+## Next Steps
+
+- **GPU support** — eventually, for large tensor operations. WebGPU is a target.
+- **Toy physics language** — I've been sketching ideas for a language optimized for numerical physics (tentatively called Éclat). It would use dimal as the foundation. No timeline yet; this is a long-term experiment.
 
 ---
 
-## Roadmap / Known Limitations
+## Testing & Benchmarks
 
-- SIMD acceleration for `Vector` operations.
+```sh
-- Some paths may still fall back to runtime computation — optimization ongoing.
+zig build test       # Run all unit tests
-- More test coverage.
+zig build benchmark  # Run performance benchmarks
+```
 
 ---
 

build.zig

@@ -4,20 +4,27 @@ pub fn build(b: *std.Build) void {
     const target = b.standardTargetOptions(.{});
     const optimize = b.standardOptimizeOption(.{});
 
-    // 1. Define the module so other projects can import it
+    const zig_wgpu = b.dependency("zig_wgpu", .{
-    _ = b.addModule("dimal", .{
+        .target = target,
-        .root_source_file = b.path("src/main.zig"),
+        .optimize = optimize,
     });
 
+    // 1. Define the module so other projects can import it
+    const mod = b.addModule("dimal", .{
+        .root_source_file = b.path("src/lib.zig"),
+    });
+    mod.addImport("gpu", zig_wgpu.module("zig-wgpu"));
+
     const exe_tests = b.addTest(.{
         .root_module = b.createModule(.{
-            .root_source_file = b.path("src/main.zig"),
+            .root_source_file = b.path("src/test.zig"),
             .target = target,
             .optimize = optimize,
         }),
         .test_runner = .{ .path = b.path("test_runner.zig"), .mode = .simple },
     });
 
+    exe_tests.root_module.addImport("gpu", zig_wgpu.module("zig-wgpu"));
     const run_exe_tests = b.addRunArtifact(exe_tests);
     const test_step = b.step("test", "Run tests");
     test_step.dependOn(&run_exe_tests.step);
@@ -30,6 +37,7 @@ pub fn build(b: *std.Build) void {
             .imports = &.{},
         }),
     });
+    bench_exe.root_module.addImport("gpu", zig_wgpu.module("zig-wgpu"));
 
     b.installArtifact(bench_exe);
 

build.zig.zon

@@ -1,9 +1,14 @@
 .{
     .name = .dimal,
-    .version = "0.1.0",
+    .version = "0.3.0",
     .fingerprint = 0x9453b1ff1e52d858,
     .minimum_zig_version = "0.16.0",
-    .dependencies = .{},
+    .dependencies = .{
+        .zig_wgpu = .{
+            .url = "git+https://git.bouvais.lu/adrien/zig-wgpu?ref=0.2.2#5f8da0940d77c40eacd39c268d09acbeaea0b2a5",
+            .hash = "zig_wgpu-0.2.0-xsLAy2-s0QPNwR2QNd8ZX2kWiVfV5oB92N3ga1V1Uwpu",
+        },
+    },
     .paths = .{
         "build.zig",
         "build.zig.zon",

docs/index.md

@@ -1,26 +0,0 @@
-# Welcome to My Project
-
-This is a static site hosted via **Gitea Actions** and **Garage S3 Storage**.
-
-!!! info "Status"
-    The deployment pipeline is currently **Active**. 
-    Updates to the `main` branch are pushed automatically.
-
-## Quick Start
-
-To replicate this setup, you need:
-1. **Traefik** as the reverse proxy.
-2. **Garage** for S3-compatible web hosting.
-3. **Gitea** for version control and CI.
-
-### Deployment Details
-| Component | Technology |
-| :--- | :--- |
-| **Engine** | MkDocs Material |
-| **Hosting** | Garage S3 |
-| **Routing** | Traefik |
-
----
-
-## Contact
-If you have questions, reach out via the Gitea instance.

docs/wishlist.md

@@ -1,3 +0,0 @@
-## GPU support with WebGPU
-
-Example: https://github.com/seyhajin/webgpu-wasm-zig

mkdocs.yml

@@ -1,48 +0,0 @@
-site_name: Bouvais Docs
-site_url: https://zig-dimal.bouvais.lu
-site_description: A minimal technical documentation site.
-site_author: Adrien Bouvais
-
-theme:
-  name: material
-  language: en
-  # Color palette with auto light/dark mode
-  palette:
-    - media: "(prefers-color-scheme: light)"
-      scheme: default
-      primary: indigo
-      accent: indigo
-      toggle:
-        icon: material/brightness-7
-        name: Switch to dark mode
-    - media: "(prefers-color-scheme: dark)"
-      scheme: slate
-      primary: indigo
-      accent: indigo
-      toggle:
-        icon: material/brightness-4
-        name: Switch to light mode
-  
-  features:
-    - navigation.sections
-    - navigation.top
-    - content.code.copy
-    - content.code.annotate
-
-# Minimal plugins
-plugins:
-  - search
-
-# Your single page
-nav:
-  - Home: index.md
-
-# Extensions to make your markdown look better
-markdown_extensions:
-  - admonition
-  - pymdownx.details
-  - pymdownx.superfences
-  - pymdownx.highlight:
-      anchor_linenums: true
-  - pymdownx.inlinehilite
-  - attr_list

src/Base.zig

@@ -3,34 +3,39 @@ const std = @import("std");
 // Adjust these imports to match your actual file names
 const Dimensions = @import("Dimensions.zig");
 const Scales = @import("Scales.zig");
-const Scalar = @import("Quantity.zig").Scalar;
+const Tensor = @import("TensorStatic.zig").Tensor;
 
 fn PhysicalConstant(comptime d: Dimensions.ArgOpts, comptime val: f64, comptime s: Scales.ArgOpts) type {
     return struct {
-        const dims = Dimensions.init(d);
+        pub const dims = Dimensions.init(d);
-        const scales = Scales.init(s);
+        pub const scales = Scales.init(s);
 
         /// Instantiates the constant into a specific numeric type.
-        pub fn Of(comptime T: type) Scalar(T, d, s) {
+        pub fn Of(comptime T: type) Tensor(T, d, s, &.{1}) {
-            return .{ .data = @splat(@as(T, @floatCast(val))) };
+            const casted_val: T = switch (@typeInfo(T)) {
+                .float => @floatCast(val),
+                .int => @intFromFloat(val),
+                else => @compileError("Unsupported type for PhysicalConstant"),
+            };
+            return Tensor(T, d, s, &.{1}).splat(casted_val);
         }
     };
 }
 
 fn BaseScalar(comptime d: Dimensions.ArgOpts) type {
     return struct {
-        const dims = Dimensions.init(d);
+        pub const dims = Dimensions.init(d);
 
         /// Creates a Scalar of this dimension using default scales.
-        /// Example: const V = Quantities.Velocity.Base(f32);
+        /// Example: const V = Quantities.Velocity.Of(f32);
         pub fn Of(comptime T: type) type {
-            return Scalar(T, d, .{});
+            return Tensor(T, d, .{}, &.{1});
         }
 
         /// Creates a Scalar of this dimension using custom scales.
-        /// Example: const Kmh = Quantities.Velocity.Scaled(f32, Scales.init(.{ .L = .k, .T = .hour }));
+        /// Example: const Kmh = Quantities.Velocity.Scaled(f32, .{ .L = .k, .T = .hour });
         pub fn Scaled(comptime T: type, comptime s: Scales.ArgOpts) type {
-            return Scalar(T, d, s);
+            return Tensor(T, d, s, &.{1});
         }
     };
 }
@@ -55,7 +60,7 @@ pub const Constants = struct {
     /// Newtonian constant of gravitation (G) [m³⋅kg⁻¹⋅s⁻²]
     pub const Gravitational = PhysicalConstant(.{ .M = -1, .L = 3, .T = -2 }, 6.67430e-11, .{ .M = .k });
 
-    /// Stefan–Boltzmann constant (σ) [W⋅m⁻²⋅K⁻⁴ = kg⋅s⁻³⋅K⁻⁴]
+    /// Stefan-Boltzmann constant () [W⋅m⁻²⋅K⁻⁴ = kg⋅s⁻³⋅K⁻⁴]
     pub const StefanBoltzmann = PhysicalConstant(.{ .M = 1, .T = -3, .Tp = -4 }, 5.670374419e-8, .{ .M = .k });
 
     /// Elementary charge (e) [C = A⋅s]
@@ -76,7 +81,7 @@ pub const Constants = struct {
     /// Neutron mass (m_n) [kg]
     pub const NeutronMass = PhysicalConstant(.{ .M = 1 }, 1.67492750056e-27, .{ .M = .k });
 
-    /// Fine-structure constant (α) [Dimensionless]
+    /// Fine-structure constant () [Dimensionless]
     pub const FineStructure = PhysicalConstant(.{}, 0.0072973525643, .{});
 
     /// Avogadro constant (N_A) [mol⁻¹]
@@ -107,7 +112,7 @@ pub const ElectricCapacitance = BaseScalar(.{ .T = 4, .L = -2, .M = -1, .I = 2 }
 pub const ElectricImpedance = ElectricResistance;
 pub const MagneticFlux = BaseScalar(.{ .M = 1, .L = 2, .T = -2, .I = -1 });
 pub const MagneticDensity = BaseScalar(.{ .M = 1, .T = -2, .I = -1 });
-pub const MagneticStrength = BaseScalar(.{ .L = -1, .I = 1 }); // Fixed typo from MagneticStrengh
+pub const MagneticStrength = BaseScalar(.{ .L = -1, .I = 1 });
 pub const MagneticMoment = BaseScalar(.{ .L = 2, .I = 1 });
 
 // ==========================================
@@ -140,7 +145,7 @@ pub const ThermalHeat = Energy;
 pub const ThermalWork = Energy;
 pub const ThermalCapacity = BaseScalar(.{ .M = 1, .L = 2, .T = -2, .Tr = -1 });
 pub const ThermalCapacityPerMass = BaseScalar(.{ .L = 2, .T = -2, .Tr = -1 });
-pub const ThermalFluxDensity = BaseScalar(.{ .M = 1, .T = -3 }); // Fixed typo from ThermalluxDensity
+pub const ThermalFluxDensity = BaseScalar(.{ .M = 1, .T = -3 });
 pub const ThermalConductance = BaseScalar(.{ .M = 1, .L = 2, .T = -3, .Tr = -1 });
 pub const ThermalConductivity = BaseScalar(.{ .M = 1, .L = 1, .T = -3, .Tr = -1 });
 pub const ThermalResistance = BaseScalar(.{ .M = -1, .L = -2, .T = 3, .Tr = 1 });
@@ -152,20 +157,24 @@ pub const ThermalEntropy = BaseScalar(.{ .M = 1, .L = 2, .T = -2, .Tr = -1 });
 // ==========================================
 pub const Frequency = BaseScalar(.{ .T = -1 });
 pub const Viscosity = BaseScalar(.{ .M = 1, .L = -1, .T = -1 });
-pub const SurfaceTension = BaseScalar(.{ .M = 1, .T = -2 }); // Corrected from MT-2a
+pub const SurfaceTension = BaseScalar(.{ .M = 1, .T = -2 });
+
+// ==========================================
+// Tests
+// ==========================================
 
 test "BaseQuantities - Core dimensions instantiation" {
     // Basic types via generic wrappers
     const M = Meter.Of(f32);
     const distance = M.splat(100);
-    try std.testing.expectEqual(100.0, distance.value());
+    try std.testing.expectEqual(100.0, distance.data[0]);
     try std.testing.expectEqual(1, M.dims.get(.L));
     try std.testing.expectEqual(0, M.dims.get(.T));
 
     // Test specific scale variants
     const Kmh = Speed.Scaled(f32, .{ .L = .k, .T = .hour });
     const speed = Kmh.splat(120);
-    try std.testing.expectEqual(120.0, speed.value());
+    try std.testing.expectEqual(120.0, speed.data[0]);
     try std.testing.expectEqual(.k, @TypeOf(speed).scales.get(.L));
     try std.testing.expectEqual(.hour, @TypeOf(speed).scales.get(.T));
 }
@@ -176,13 +185,13 @@ test "BaseQuantities - Kinematics equations" {
 
     // Velocity = Distance / Time
     const v = d.div(t);
-    try std.testing.expectEqual(25.0, v.value());
+    try std.testing.expectEqual(25.0, v.data[0]);
-    try std.testing.expect(Speed.dims.eql(@TypeOf(v).dims));
+    try comptime std.testing.expect(Speed.dims.eql(@TypeOf(v).dims));
 
     // Acceleration = Velocity / Time
     const a = v.div(t);
-    try std.testing.expectEqual(12.5, a.value());
+    try std.testing.expectEqual(12.5, a.data[0]);
-    try std.testing.expect(Acceleration.dims.eql(@TypeOf(a).dims));
+    try comptime std.testing.expect(Acceleration.dims.eql(@TypeOf(a).dims));
 }
 
 test "BaseQuantities - Dynamics (Force and Work)" {
@@ -193,14 +202,14 @@ test "BaseQuantities - Dynamics (Force and Work)" {
 
     // Force = mass * acceleration
     const f = m.mul(a);
-    try std.testing.expectEqual(98, f.value());
+    try std.testing.expectEqual(98, f.data[0]);
-    try std.testing.expect(Force.dims.eql(@TypeOf(f).dims));
+    try comptime std.testing.expect(Force.dims.eql(@TypeOf(f).dims));
 
     // Energy (Work) = Force * distance
     const distance = Meter.Of(f32).splat(5.0);
     const energy = f.mul(distance);
-    try std.testing.expectEqual(490, energy.value());
+    try std.testing.expectEqual(490, energy.data[0]);
-    try std.testing.expect(Energy.dims.eql(@TypeOf(energy).dims));
+    try comptime std.testing.expect(Energy.dims.eql(@TypeOf(energy).dims));
 }
 
 test "BaseQuantities - Electric combinations" {
@@ -209,26 +218,26 @@ test "BaseQuantities - Electric combinations" {
 
     // Charge = Current * time
     const charge = current.mul(time);
-    try std.testing.expectEqual(6.0, charge.value());
+    try std.testing.expectEqual(6.0, charge.data[0]);
-    try std.testing.expect(ElectricCharge.dims.eql(@TypeOf(charge).dims));
+    try comptime std.testing.expect(ElectricCharge.dims.eql(@TypeOf(charge).dims));
 }
 
 test "Constants - Initialization and dimension checks" {
     // Speed of Light
     const c = Constants.SpeedOfLight.Of(f64);
-    try std.testing.expectEqual(299792458.0, c.value());
+    try std.testing.expectEqual(299792458.0, c.data[0]);
     try std.testing.expectEqual(1, @TypeOf(c).dims.get(.L));
     try std.testing.expectEqual(-1, @TypeOf(c).dims.get(.T));
 
     // Electron Mass (verifying scale as well)
     const me = Constants.ElectronMass.Of(f64);
-    try std.testing.expectEqual(9.1093837139e-31, me.value());
+    try std.testing.expectEqual(9.1093837139e-31, me.data[0]);
     try std.testing.expectEqual(1, @TypeOf(me).dims.get(.M));
     try std.testing.expectEqual(.k, @TypeOf(me).scales.get(.M)); // Should be scaled to kg
 
     // Boltzmann Constant (Complex derived dimensions)
     const kb = Constants.Boltzmann.Of(f64);
-    try std.testing.expectEqual(1.380649e-23, kb.value());
+    try std.testing.expectEqual(1.380649e-23, kb.data[0]);
     try std.testing.expectEqual(1, @TypeOf(kb).dims.get(.M));
     try std.testing.expectEqual(2, @TypeOf(kb).dims.get(.L));
     try std.testing.expectEqual(-2, @TypeOf(kb).dims.get(.T));
@@ -237,7 +246,7 @@ test "Constants - Initialization and dimension checks" {
 
     // Vacuum Permittivity
     const eps0 = Constants.VacuumPermittivity.Of(f64);
-    try std.testing.expectEqual(8.8541878188e-12, eps0.value());
+    try std.testing.expectEqual(8.8541878188e-12, eps0.data[0]);
     try std.testing.expectEqual(-1, @TypeOf(eps0).dims.get(.M));
     try std.testing.expectEqual(-3, @TypeOf(eps0).dims.get(.L));
     try std.testing.expectEqual(4, @TypeOf(eps0).dims.get(.T));
@@ -245,7 +254,7 @@ test "Constants - Initialization and dimension checks" {
 
     // Fine Structure Constant (Dimensionless)
     const alpha = Constants.FineStructure.Of(f64);
-    try std.testing.expectEqual(0.0072973525643, alpha.value());
+    try std.testing.expectEqual(0.0072973525643, alpha.data[0]);
     try std.testing.expectEqual(0, @TypeOf(alpha).dims.get(.M));
     try std.testing.expectEqual(0, @TypeOf(alpha).dims.get(.L));
 }

src/Dimensions.zig

@@ -1,13 +1,13 @@
 const std = @import("std");
 
 pub const ArgOpts = struct {
-    L: comptime_int = 0,
+    L: isize = 0,
-    M: comptime_int = 0,
+    M: isize = 0,
-    T: comptime_int = 0,
+    T: isize = 0,
-    I: comptime_int = 0,
+    I: isize = 0,
-    Tp: comptime_int = 0,
+    Tp: isize = 0,
-    N: comptime_int = 0,
+    N: isize = 0,
-    J: comptime_int = 0,
+    J: isize = 0,
 };
 
 pub const Dimension = enum {
@@ -49,81 +49,81 @@ data: std.EnumArray(Dimension, comptime_int),
 
 /// Create a `Dimensions` from a struct literal, e.g. `.{ .L = 1, .T = -1 }`.
 /// Unspecified dimensions default to 0.
-pub fn init(comptime init_val: ArgOpts) Self {
+pub fn init(init_val: ArgOpts) Self {
     var s = Self{ .data = std.EnumArray(Dimension, comptime_int).initFill(0) };
-    inline for (std.meta.fields(@TypeOf(init_val))) |f|
+    for (std.meta.fields(@TypeOf(init_val))) |f|
         s.data.set(@field(Dimension, f.name), @field(init_val, f.name));
     return s;
 }
 
-pub fn initFill(comptime val: comptime_int) Self {
+pub fn initFill(val: comptime_int) Self {
     return .{ .data = std.EnumArray(Dimension, comptime_int).initFill(val) };
 }
 
-pub fn get(comptime self: Self, comptime key: Dimension) comptime_int {
+pub fn get(self: Self, key: Dimension) comptime_int {
     return self.data.get(key);
 }
 
-pub fn set(comptime self: *Self, comptime key: Dimension, comptime val: i8) void {
+pub fn set(self: *Self, key: Dimension, val: i8) void {
     self.data.set(key, val);
 }
 
 pub fn argsOpt(self: Self) ArgOpts {
     var args: ArgOpts = undefined;
-    inline for (std.enums.values(Dimension)) |d|
+    for (std.enums.values(Dimension)) |d|
         @field(args, @tagName(d)) = self.get(d);
     return args;
 }
 
 /// Add exponents component-wise. Used internally by `mul`.
-pub fn add(comptime a: Self, comptime b: Self) Self {
+pub fn add(a: Self, b: Self) Self {
     var result = Self.initFill(0);
-    inline for (std.enums.values(Dimension)) |d|
+    for (std.enums.values(Dimension)) |d|
         result.set(d, a.get(d) + b.get(d));
     return result;
 }
 
 /// Subtract exponents component-wise. Used internally by `div`.
-pub fn sub(comptime a: Self, comptime b: Self) Self {
+pub fn sub(a: Self, b: Self) Self {
     var result = Self.initFill(0);
-    inline for (std.enums.values(Dimension)) |d|
+    for (std.enums.values(Dimension)) |d|
         result.set(d, a.get(d) - b.get(d));
     return result;
 }
 
 /// Multiply exponents by a scalar integer. Used internally by `pow` in Scalar.
-pub fn scale(comptime a: Self, comptime exp: comptime_int) Self {
+pub fn scale(a: Self, exp: comptime_int) Self {
     var result = Self.initFill(0);
-    inline for (std.enums.values(Dimension)) |d|
+    for (std.enums.values(Dimension)) |d|
         result.set(d, a.get(d) * exp);
     return result;
 }
 
-pub fn div(comptime a: Self, comptime exp: comptime_int) Self {
+pub fn div(a: Self, exp: comptime_int) Self {
     var result = Self.initFill(0);
-    inline for (std.enums.values(Dimension)) |d|
+    for (std.enums.values(Dimension)) |d|
         result.set(d, a.get(d) / exp);
     return result;
 }
 
 /// Returns true if every dimension exponent is equal. Used to enforce type compatibility in `add`, `sub`, `to`.
-pub fn eql(comptime a: Self, comptime b: Self) bool {
+pub fn eql(a: Self, b: Self) bool {
-    inline for (std.enums.values(Dimension)) |d|
+    for (std.enums.values(Dimension)) |d|
         if (a.get(d) != b.get(d)) return false;
     return true;
 }
 
-pub fn isSquare(comptime a: Self) bool {
+pub fn isSquare(a: Self) bool {
-    inline for (std.enums.values(Dimension)) |d|
+    for (std.enums.values(Dimension)) |d|
         if (a.get(d) % 2 != 0) return false;
     return true;
 }
 
-pub fn str(comptime a: Self) []const u8 {
+pub fn str(a: Self) []const u8 {
     var out: []const u8 = "";
     const dims = std.enums.values(Dimension);
 
-    inline for (dims) |d| {
+    for (dims) |d| {
         const val = a.get(d);
         if (val != 0) {
             out = out ++ @tagName(d) ++ std.fmt.comptimePrint("{d}", .{val});

src/Scales.zig

@@ -1,5 +1,4 @@
 const std = @import("std");
-const hlp = @import("helper.zig");
 const Dimensions = @import("Dimensions.zig");
 const Dimension = @import("Dimensions.zig").Dimension;
 
@@ -56,35 +55,36 @@ pub const UnitScale = enum(isize) {
     // Undefined
     _,
 
-    pub inline fn str(self: @This()) []const u8 {
+    pub fn str(self: @This()) []const u8 {
         var buf: [16]u8 = undefined;
         return switch (self) {
-            inline .none => "",
+            .none => "",
-            inline .P, .T, .G, .M, .k, .h, .da, .d, .c, .m, .u, .n, .p, .f, .min, .hour, .year, .inch, .ft, .yd, .mi, .oz, .lb, .st => @tagName(self),
+            .P, .T, .G, .M, .k, .h, .da, .d, .c, .m, .u, .n, .p, .f, .min, .year, .inch, .ft, .yd, .mi, .oz, .lb, .st => @tagName(self),
+            .hour => "h",
             else => std.fmt.bufPrint(&buf, "[{d}]", .{@intFromEnum(self)}) catch "[]", // This cannot be inline because of non exhaustive enum, but that's ok, it is just str, not calculation
         };
     }
 
-    pub inline fn getFactor(self: @This()) comptime_float {
+    pub fn getFactor(self: @This()) comptime_float {
-        return comptime switch (self) {
+        return switch (self) {
             // Standard SI Exponents
-            inline .P, .T, .G, .M, .k, .h, .da, .none, .d, .c, .m, .u, .n, .p, .f => std.math.pow(f64, 10.0, @floatFromInt(@intFromEnum(self))),
+            .P, .T, .G, .M, .k, .h, .da, .none, .d, .c, .m, .u, .n, .p, .f => std.math.pow(f64, 10.0, @floatFromInt(@intFromEnum(self))),
 
             // Time Factors
-            inline .min, .hour, .year => @floatFromInt(@intFromEnum(self)),
+            .min, .hour, .year => @floatFromInt(@intFromEnum(self)),
 
             // Imperial Length (metres)
-            inline .inch => 0.0254,
+            .inch => 0.0254,
-            inline .ft => 0.3048,
+            .ft => 0.3048,
-            inline .yd => 0.9144,
+            .yd => 0.9144,
-            inline .mi => 1609.344,
+            .mi => 1609.344,
 
             // Imperial Mass (grams — base unit for M is gram, i.e. .none = 1 g)
-            inline .oz => 28.3495231,
+            .oz => 28.3495231,
-            inline .lb => 453.59237,
+            .lb => 453.59237,
-            inline .st => 6350.29318,
+            .st => 6350.29318,
 
-            inline else => @floatFromInt(@intFromEnum(self)),
+            else => @floatFromInt(@intFromEnum(self)),
         };
     }
 };
@@ -98,40 +98,46 @@ data: std.EnumArray(Dimension, UnitScale),
 /// Unspecified dimensions default to `.none` (factor 1).
 pub fn init(comptime init_val: ArgOpts) Self {
     comptime var s = Self{ .data = std.EnumArray(Dimension, UnitScale).initFill(.none) };
-    inline for (std.meta.fields(@TypeOf(init_val))) |f| {
+    for (std.meta.fields(@TypeOf(init_val))) |f| {
-        if (comptime hlp.isInt(@TypeOf(@field(init_val, f.name))))
+        if (comptime @typeInfo(@TypeOf(@field(init_val, f.name))) == .comptime_int)
             s.data.set(@field(Dimension, f.name), @enumFromInt(@field(init_val, f.name)))
         else
             s.data.set(@field(Dimension, f.name), @field(init_val, f.name));
     }
-    return s;
+    return comptime s;
 }
 
-pub fn initFill(comptime val: UnitScale) Self {
+pub fn initFill(val: UnitScale) Self {
-    return comptime .{ .data = std.EnumArray(Dimension, UnitScale).initFill(val) };
+    return .{ .data = std.EnumArray(Dimension, UnitScale).initFill(val) };
 }
 
-pub fn get(comptime self: Self, comptime key: Dimension) UnitScale {
+pub fn get(self: Self, key: Dimension) UnitScale {
-    return comptime self.data.get(key);
+    return self.data.get(key);
 }
 
-pub fn set(comptime self: *Self, comptime key: Dimension, comptime val: UnitScale) void {
+pub fn set(self: *Self, key: Dimension, val: UnitScale) void {
-    comptime self.data.set(key, val);
+    self.data.set(key, val);
+}
+
+pub fn eql(self: Self, other: Self) bool {
+    for (self.data.values, other.data.values) |l, r|
+        if (l != r) return false;
+    return true;
 }
 
 pub fn argsOpt(self: Self) ArgOpts {
     var args: ArgOpts = undefined;
-    inline for (std.enums.values(Dimension)) |d|
+    for (std.enums.values(Dimension)) |d|
         @field(args, @tagName(d)) = self.get(d);
     return args;
 }
 
 /// Compute the combined scale factor for a given dimension signature.
 /// Each dimension's prefix is raised to its exponent and multiplied together.
-pub inline fn getFactor(comptime s: Self, comptime d: Dimensions) comptime_float {
+pub fn getFactor(s: Self, d: Dimensions) comptime_float {
     var factor: f64 = 1.0;
     for (std.enums.values(Dimension)) |dim| {
-        const power = comptime d.get(dim);
+        const power = d.get(dim);
         if (power == 0) continue;
 
         const base = s.get(dim).getFactor();
@@ -145,5 +151,5 @@ pub inline fn getFactor(comptime s: Self, comptime d: Dimensions) comptime_float
                 factor /= base;
         }
     }
-    return comptime factor;
+    return factor;
 }

src/UnitParser.zig

@@ -0,0 +1,145 @@
+const std = @import("std");
+const Dimensions = @import("Dimensions.zig");
+const Scales = @import("Scales.zig");
+
+/// A container returning the separated arguments needed to construct a Tensor.
+pub const ParsedUnit = struct {
+    dims: Dimensions.ArgOpts = .{},
+    scales: Scales.ArgOpts = .{},
+};
+
+pub const UnitParseError = error{
+    UnknownBaseUnit,
+    UnknownPrefix,
+    InvalidExponent,
+    EmptyStr,
+};
+
+/// Parses strings like "km/s^2", "m", "kg*m/s^2", "1/min".
+/// Evaluates entirely at comptime.
+pub fn parseUnit(str: []const u8) !ParsedUnit {
+    if (str.len == 0) return UnitParseError.EmptyStr;
+
+    var parsed: ParsedUnit = .{ .dims = .{}, .scales = .{} };
+
+    // We need to track if we are after a '/' to flip exponents to negative
+    var is_denominator = false;
+
+    // Manual iteration to handle '/' properly
+    var cursor: usize = 0;
+    while (cursor < str.len) {
+        // Find the next segment
+        const segment_start = cursor;
+        while (cursor < str.len and str[cursor] != '/' and str[cursor] != '.' and str[cursor] != '*') : (cursor += 1) {}
+        const segment = str[segment_start..cursor];
+
+        if (segment.len > 0) {
+            try parseSegment(segment, &parsed, is_denominator);
+        }
+
+        if (cursor < str.len) {
+            if (str[cursor] == '/') {
+                is_denominator = true;
+            }
+            cursor += 1; // skip the separator
+        }
+    }
+
+    return parsed;
+}
+
+fn parseSegment(segment: []const u8, parsed: *ParsedUnit, is_denominator: bool) !void {
+    var scale: Scales.UnitScale = .none;
+    var found_scale = false;
+    var active_dim: ?Dimensions.Dimension = null;
+
+    // 1. Try to find a Scale + Dimension pair (e.g., "mm", "km")
+    inline for (std.enums.values(Scales.UnitScale)) |sca| {
+        const s_str = sca.str();
+        if (s_str.len > 0 and std.mem.startsWith(u8, segment, s_str)) {
+            // Check if it's a "Unit-as-Scale" (hour, min) or a prefix (k, m, c)
+            switch (sca) {
+                .hour, .min, .year => {
+                    // These are dimensions themselves (Time)
+                    if (segment.len == s_str.len or (segment.len > s_str.len and (segment[s_str.len] == '^' or (segment[s_str.len] >= '0' and segment[s_str.len] <= '9')))) {
+                        scale = sca;
+                        active_dim = .T;
+                        found_scale = true;
+                    }
+                },
+                else => {
+                    // Standard prefixes: Must be followed by a valid dimension unit
+                    inline for (std.enums.values(Dimensions.Dimension)) |dim| {
+                        if (std.mem.startsWith(u8, segment[s_str.len..], dim.unit())) {
+                            scale = sca;
+                            active_dim = dim;
+                            found_scale = true;
+                            break;
+                        }
+                    }
+                },
+            }
+        }
+        if (found_scale) break;
+    }
+
+    // 2. If no scale prefix was found, try identifying as a pure Dimension (e.g., "m", "s")
+    if (!found_scale) {
+        inline for (std.enums.values(Dimensions.Dimension)) |dim| {
+            if (std.mem.startsWith(u8, segment, dim.unit())) {
+                active_dim = dim;
+                break;
+            }
+        }
+    }
+
+    const dimen = active_dim orelse return UnitParseError.UnknownBaseUnit;
+
+    // 3. Determine where the exponent starts
+    // If it was a Time Scale (like 'h'), the exponent starts after 'h'
+    // If it was a Prefix + Dim (like 'km'), it starts after 'km'
+    const unit_part_len = if (found_scale)
+        (if (scale == .hour or scale == .min or scale == .year) scale.str().len else scale.str().len + dimen.unit().len)
+    else
+        dimen.unit().len;
+
+    const expo_str = segment[unit_part_len..];
+
+    // 4. Parse Exponent
+    var expo: i32 = 1;
+    if (expo_str.len > 0) {
+        const cleaned_expo = if (expo_str[0] == '^') expo_str[1..] else expo_str;
+        expo = std.fmt.parseInt(i32, cleaned_expo, 10) catch return UnitParseError.InvalidExponent;
+    }
+
+    if (is_denominator) expo *= -1;
+
+    // 5. Assign to struct
+    inline for (std.meta.fields(Dimensions.ArgOpts)) |f| {
+        if (std.mem.eql(u8, f.name, @tagName(dimen))) {
+            @field(parsed.dims, f.name) += expo;
+            @field(parsed.scales, f.name) = scale;
+        }
+    }
+}
+
+inline fn testParser(
+    comptime str: []const u8,
+    comptime expected_dims: Dimensions.ArgOpts,
+    comptime expected_scales: Scales.ArgOpts,
+) !void {
+    const unit = comptime try parseUnit(str);
+    if (comptime !Dimensions.init(expected_dims).eql(Dimensions.init(unit.dims))) return error.WrongDims;
+    if (comptime !Scales.init(expected_scales).eql(Scales.init(unit.scales))) return error.WrongScales;
+}
+
+test "parseUnit" {
+    @setEvalBranchQuota(10000);
+    try testParser("m", .{ .L = 1 }, .{});
+    try testParser("s", .{ .T = 1 }, .{});
+    try testParser("mm", .{ .L = 1 }, .{ .L = .m });
+    try testParser("m/s", .{ .L = 1, .T = -1 }, .{});
+    try testParser("m1/s2/kg", .{ .L = 1, .T = -2, .M = -1 }, .{ .M = .k });
+    try testParser("km/h", .{ .L = 1, .T = -1 }, .{ .L = .k, .T = .hour });
+    try testParser("m.s^-1", .{ .L = 1, .T = -1 }, .{});
+}

src/benchmark.zig

@@ -1,7 +1,6 @@
 const std = @import("std");
 const Io = std.Io;
-const Scalar = @import("Quantity.zig").Scalar;
+const Tensor = @import("Tensor.zig").Tensor;
-const Vector = @import("Quantity.zig").Vector;
 
 var io: Io = undefined;
 pub fn main(init: std.process.Init) !void {
@@ -11,25 +10,12 @@ pub fn main(init: std.process.Init) !void {
 
     io = init.io;
 
-    // try vectorSIMDvsNative(f64, &stdout_writer.interface);
-    // try stdout_writer.flush();
-    // try vectorSIMDvsNative(f32, &stdout_writer.interface);
-    // try stdout_writer.flush();
-    // try vectorSIMDvsNative(i32, &stdout_writer.interface);
-    // try stdout_writer.flush();
-    // try vectorSIMDvsNative(i64, &stdout_writer.interface);
-    // try stdout_writer.flush();
-    // try vectorSIMDvsNative(i128, &stdout_writer.interface);
-    // try stdout_writer.flush();
-
     try bench_Scalar(&stdout_writer.interface);
     try stdout_writer.flush();
-    try bench_vsNative(&stdout_writer.interface);
-    try stdout_writer.flush();
-    try bench_crossTypeVsNative(&stdout_writer.interface);
-    try stdout_writer.flush();
     try bench_Vector(&stdout_writer.interface);
     try stdout_writer.flush();
+    try bench_HighDimTensor(&stdout_writer.interface);
+    try stdout_writer.flush();
 }
 
 fn getTime() Io.Timestamp {
@@ -97,9 +83,9 @@ fn bench_Scalar(writer: *std.Io.Writer) !void {
 
     comptime var tidx: usize = 0;
     inline for (Types, TNames) |T, tname| {
-        const M = Scalar(T, .{ .L = 1 }, .{});
+        const M = Tensor(T, .{ .L = 1 }, .{}, &.{1});
-        const KM = Scalar(T, .{ .L = 1 }, .{ .L = .k });
+        const KM = Tensor(T, .{ .L = 1 }, .{ .L = .k }, &.{1});
-        const S = Scalar(T, .{ .T = 1 }, .{});
+        const S = Tensor(T, .{ .T = 1 }, .{}, &.{1});
 
         inline for (Ops, 0..) |op_name, oidx| {
             var samples: [SAMPLES]f64 = undefined;
@@ -127,7 +113,7 @@ fn bench_Scalar(writer: *std.Io.Writer) !void {
                             else if (comptime std.mem.eql(u8, op_name, "gt"))
                                 (M.splat(getVal(T, i, 63))).gt(M.splat(getVal(T, i +% 3, 63)))
                             else
-                                (M.splat(getVal(T, i, 63))).mul(3);
+                                (M.splat(getVal(T, i, 63))).mul(M.splat(3));
                         },
                     );
                 }
@@ -168,222 +154,6 @@ fn bench_Scalar(writer: *std.Io.Writer) !void {
     try writer.print("└──────────────┴───────┴───────┴───────┴───────┴───────┴───────┴───────┘\n", .{});
 }
 
-fn bench_vsNative(writer: *std.Io.Writer) !void {
-    const ITERS: usize = 100_000;
-    const SAMPLES: usize = 5;
-
-    const getValT = struct {
-        fn f(comptime TT: type, i: usize) TT {
-            const v = (i % 100) + 1;
-            return if (comptime @typeInfo(TT) == .float) @floatFromInt(v) else @intCast(v);
-        }
-    }.f;
-
-    const Types = .{ i32, i64, i128, f32, f64 };
-    const TNames = .{ "i32", "i64", "i128", "f32", "f64" };
-    // Expanded Ops to match bench_Scalar
-    const Ops = .{ "add", "sub", "mul", "div", "abs", "eq", "gt" };
-
-    try writer.print(
-        \\
-        \\ Scalar vs Native Overhead Analysis
-        \\
-        \\┌───────────┬──────┬───────────┬───────────┬───────────┐
-        \\│ Operation │ Type │ Native    │ Scalar    │ Slowdown  │
-        \\├───────────┼──────┼───────────┼───────────┼───────────┤
-        \\
-    , .{});
-
-    inline for (Ops, 0..) |op_name, j| {
-        inline for (Types, 0..) |T, tidx| {
-            var native_total_ns: f64 = 0;
-            var quantity_total_ns: f64 = 0;
-
-            const M = Scalar(T, .{ .L = 1 }, .{});
-            const S = Scalar(T, .{ .T = 1 }, .{});
-
-            std.mem.doNotOptimizeAway({
-                for (0..SAMPLES) |_| {
-                    // --- 1. Benchmark Native ---
-                    const n_start = getTime();
-                    for (0..ITERS) |i| {
-                        const a = getValT(T, i);
-                        const b = getValT(T, 2);
-
-                        // Native logic branch
-                        _ = if (comptime std.mem.eql(u8, op_name, "add"))
-                            a + b
-                        else if (comptime std.mem.eql(u8, op_name, "sub"))
-                            a - b
-                        else if (comptime std.mem.eql(u8, op_name, "mul"))
-                            a * b
-                        else if (comptime std.mem.eql(u8, op_name, "div"))
-                            if (comptime @typeInfo(T) == .int) @divTrunc(a, b) else a / b
-                        else if (comptime std.mem.eql(u8, op_name, "abs"))
-                            if (comptime @typeInfo(T) == .int) @abs(a) else @as(T, @abs(a))
-                        else if (comptime std.mem.eql(u8, op_name, "eq"))
-                            a == b
-                        else if (comptime std.mem.eql(u8, op_name, "gt"))
-                            a > b
-                        else
-                            unreachable;
-                    }
-                    const n_end = getTime();
-                    native_total_ns += @as(f64, @floatFromInt(n_start.durationTo(n_end).toNanoseconds()));
-
-                    // --- 2. Benchmark Scalar ---
-                    const q_start = getTime();
-                    for (0..ITERS) |i| {
-                        const qa = M.splat(getValT(T, i));
-                        const qb = if (comptime std.mem.eql(u8, op_name, "div")) S.splat(getValT(T, 2)) else M.splat(getValT(T, 2));
-
-                        // Scalar logic branch
-                        _ = if (comptime std.mem.eql(u8, op_name, "add"))
-                            qa.add(qb)
-                        else if (comptime std.mem.eql(u8, op_name, "sub"))
-                            qa.sub(qb)
-                        else if (comptime std.mem.eql(u8, op_name, "mul"))
-                            qa.mul(qb)
-                        else if (comptime std.mem.eql(u8, op_name, "div"))
-                            qa.div(qb)
-                        else if (comptime std.mem.eql(u8, op_name, "abs"))
-                            qa.abs()
-                        else if (comptime std.mem.eql(u8, op_name, "eq"))
-                            qa.eq(qb)
-                        else if (comptime std.mem.eql(u8, op_name, "gt"))
-                            qa.gt(qb)
-                        else
-                            unreachable;
-                    }
-                    const q_end = getTime();
-                    quantity_total_ns += @as(f64, @floatFromInt(q_start.durationTo(q_end).toNanoseconds()));
-                }
-            });
-
-            const avg_n = (native_total_ns / SAMPLES) / @as(f64, @floatFromInt(ITERS));
-            const avg_q = (quantity_total_ns / SAMPLES) / @as(f64, @floatFromInt(ITERS));
-            const slowdown = avg_q / avg_n;
-
-            try writer.print("│ {s:<9} │ {s:<4} │ {d:>7.2}ns │ {d:>7.2}ns │ {d:>8.2}x │\n", .{
-                op_name, TNames[tidx], avg_n, avg_q, slowdown,
-            });
-        }
-        if (j != Ops.len - 1) try writer.print("├───────────┼──────┼───────────┼───────────┼───────────┤\n", .{});
-    }
-
-    try writer.print("└───────────┴──────┴───────────┴───────────┴───────────┘\n", .{});
-}
-
-fn bench_crossTypeVsNative(writer: *std.Io.Writer) !void {
-    const ITERS: usize = 100_000;
-    const SAMPLES: usize = 5;
-
-    const getValT = struct {
-        fn f(comptime TT: type, i: usize) TT {
-            // Keep values safe and non-zero to avoid division by zero or overflows during cross-casting
-            const v = (i % 50) + 1;
-            return if (comptime @typeInfo(TT) == .float) @floatFromInt(v) else @intCast(v);
-        }
-    }.f;
-
-    // Helper for the Native baseline: explicitly casting T2 to T1 before the operation
-    const castTo = struct {
-        fn f(comptime DestT: type, comptime SrcT: type, val: SrcT) DestT {
-            if (comptime DestT == SrcT) return val;
-            const src_info = @typeInfo(SrcT);
-            const dest_info = @typeInfo(DestT);
-
-            if (dest_info == .int and src_info == .int) return @intCast(val);
-            if (dest_info == .float and src_info == .int) return @floatFromInt(val);
-            if (dest_info == .int and src_info == .float) return @intFromFloat(val);
-            if (dest_info == .float and src_info == .float) return @floatCast(val);
-            unreachable;
-        }
-    }.f;
-
-    const Types = .{ i16, i64, i128, f32, f64 };
-    const TNames = .{ "i16", "i64", "i128", "f32", "f64" };
-    const Ops = .{ "add", "mul", "div" };
-
-    try writer.print(
-        \\
-        \\ Cross-Type Overhead Analysis: Scalar vs Native
-        \\
-        \\┌─────────┬──────┬──────┬───────────┬───────────┬───────────┐
-        \\│ Op      │ T1   │ T2   │ Native    │ Scalar    │ Slowdown  │
-        \\├─────────┼──────┼──────┼───────────┼───────────┼───────────┤
-        \\
-    , .{});
-
-    inline for (Ops, 0..) |op_name, j| {
-        inline for (Types, 0..) |T1, t1_idx| {
-            inline for (Types, 0..) |T2, t2_idx| {
-                var native_total_ns: f64 = 0;
-                var quantity_total_ns: f64 = 0;
-
-                const M1 = Scalar(T1, .{ .L = 1 }, .{});
-                const M2 = Scalar(T2, .{ .L = 1 }, .{});
-                const S2 = Scalar(T2, .{ .T = 1 }, .{});
-
-                std.mem.doNotOptimizeAway({
-                    for (0..SAMPLES) |_| {
-                        // --- 1. Benchmark Native (Cast T2 to T1, then math) ---
-                        const n_start = getTime();
-                        for (0..ITERS) |i| {
-                            const a = getValT(T1, i);
-                            const b_raw = getValT(T2, 2);
-                            const b = castTo(T1, T2, b_raw);
-
-                            _ = if (comptime std.mem.eql(u8, op_name, "add"))
-                                a + b
-                            else if (comptime std.mem.eql(u8, op_name, "mul"))
-                                a * b
-                            else if (comptime @typeInfo(T1) == .int)
-                                @divTrunc(a, b)
-                            else
-                                a / b;
-                        }
-                        const n_end = getTime();
-                        native_total_ns += @as(f64, @floatFromInt(n_start.durationTo(n_end).toNanoseconds()));
-
-                        // --- 2. Benchmark Scalar ---
-                        const q_start = getTime();
-                        for (0..ITERS) |i| {
-                            const qa = M1.splat(getValT(T1, i));
-                            const qb = if (comptime std.mem.eql(u8, op_name, "div"))
-                                S2.splat(getValT(T2, 2))
-                            else
-                                M2.splat(getValT(T2, 2));
-
-                            _ = if (comptime std.mem.eql(u8, op_name, "add"))
-                                qa.add(qb)
-                            else if (comptime std.mem.eql(u8, op_name, "mul"))
-                                qa.mul(qb)
-                            else
-                                qa.div(qb);
-                        }
-                        const q_end = getTime();
-                        quantity_total_ns += @as(f64, @floatFromInt(q_start.durationTo(q_end).toNanoseconds()));
-                    }
-
-                    const avg_n = (native_total_ns / SAMPLES) / @as(f64, @floatFromInt(ITERS));
-                    const avg_q = (quantity_total_ns / SAMPLES) / @as(f64, @floatFromInt(ITERS));
-                    const slowdown = avg_q / avg_n;
-
-                    try writer.print("│ {s:<7} │ {s:<4} │ {s:<4} │ {d:>7.2}ns │ {d:>7.2}ns │ {d:>8.2}x │\n", .{
-                        op_name, TNames[t1_idx], TNames[t2_idx], avg_n, avg_q, slowdown,
-                    });
-                });
-            }
-        }
-        if (j != Ops.len - 1) {
-            try writer.print("├─────────┼──────┼──────┼───────────┼───────────┼───────────┤\n", .{});
-        }
-    }
-
-    try writer.print("└─────────┴──────┴──────┴───────────┴───────────┴───────────┘\n", .{});
-}
-
 fn bench_Vector(writer: *std.Io.Writer) !void {
     const ITERS: usize = 10_000;
     const SAMPLES: usize = 10;
@@ -422,16 +192,15 @@ fn bench_Vector(writer: *std.Io.Writer) !void {
     const TNames = .{ "i32", "i64", "i128", "f32", "f64" };
     const Lengths = .{ 1, 3, 4, 16, 100 };
     // "cross" is only valid for len=3; other cells will show "  ---  "
-    const Ops = .{ "add", "div", "mulScalar", "dot", "cross", "product", "pow", "length" };
+    const Ops = .{ "add", "div", "mulScalar", "dot", "product", "pow", "length" };
 
     inline for (Ops, 0..) |op_name, o_idx| {
         inline for (Types, TNames) |T, tname| {
             try writer.print("│ {s:<16} │ {s:<4} │", .{ op_name, tname });
 
             inline for (Lengths) |len| {
-                const Q_base = Scalar(T, .{ .L = 1 }, .{});
+                const Q_time = Tensor(T, .{ .T = 1 }, .{}, &.{1});
-                const Q_time = Scalar(T, .{ .T = 1 }, .{});
+                const V = Tensor(T, .{ .L = 1 }, .{}, &.{len});
-                const V = Vector(len, Q_base);
 
                 // cross product is only defined for len == 3
                 const is_cross = comptime std.mem.eql(u8, op_name, "cross");
@@ -455,14 +224,10 @@ fn bench_Vector(writer: *std.Io.Writer) !void {
                                 _ = v1.div(V.splat(getVal(T, i +% 2, 63)));
                             } else if (comptime std.mem.eql(u8, op_name, "mulScalar")) {
                                 const s_val = Q_time.splat(getVal(T, i +% 2, 63));
-                                _ = v1.mulScalar(s_val);
+                                _ = v1.mul(s_val);
                             } else if (comptime std.mem.eql(u8, op_name, "dot")) {
                                 const v2 = V.splat(getVal(T, i +% 5, 63));
-                                _ = v1.dot(v2);
+                                _ = v1.contract(v2, 0, 0);
-                            } else if (comptime std.mem.eql(u8, op_name, "cross")) {
-                                // len == 3 guaranteed by the guard above
-                                const v2 = V.splat(getVal(T, i +% 5, 63));
-                                _ = v1.cross(v2);
                             } else if (comptime std.mem.eql(u8, op_name, "product")) {
                                 _ = v1.product();
                             } else if (comptime std.mem.eql(u8, op_name, "pow")) {
@@ -490,61 +255,98 @@ fn bench_Vector(writer: *std.Io.Writer) !void {
     try writer.print("└──────────────────┴──────┴─────────┴─────────┴─────────┴─────────┴─────────┘\n", .{});
 }
 
-fn vectorSIMDvsNative(comptime T: type, writer: *std.Io.Writer) !void {
+fn bench_HighDimTensor(writer: *std.Io.Writer) !void {
-    const iterations: u64 = 10_000;
+    const ITERS: usize = 5_000;
-    const lens = [_]u32{ 1, 2, 3, 4, 5, 10, 100, 1_000, 10_000 };
+    const SAMPLES: usize = 5;
 
-    try writer.print("\nSIMD Speedup Analysis: {s}\n", .{@typeName(T)});
+    const getVal = struct {
-    try writer.print("┌────────────┬────────────┬────────────┬────────────┐\n", .{});
+        fn f(comptime TT: type, i: usize, comptime mask: u7) TT {
-    try writer.print("│ Vector Len │ Scalar (us)│ Vector (us)│ Speedup    │\n", .{});
+            const v: u8 = @as(u8, @truncate(i & @as(usize, mask))) + 1;
-    try writer.print("├────────────┼────────────┼────────────┼────────────┤\n", .{});
+            return if (comptime @typeInfo(TT) == .float) @floatFromInt(v) else @intCast(v);
-
-    inline for (lens) |vector_len| {
-        // --- Scalar Test ---
-        var scalar_val: T = 10;
-        const start_scalar = getTime();
-
-        var i: u64 = 0;
-        while (i < iterations * vector_len) : (i += 1) {
-            if (comptime @typeInfo(T) == .int)
-                scalar_val = scalar_val +% 1
-            else
-                scalar_val = scalar_val + 1;
         }
-        const scalar_time = start_scalar.durationTo(getTime()).toMicroseconds();
+    }.f;
 
-        // --- Vector Test ---
+    const computeStats = struct {
-        var vector_val: @Vector(vector_len, T) = @splat(20);
+        fn f(samples: []f64, iters: usize) f64 {
-        const start_vector = getTime();
+            std.mem.sort(f64, samples, {}, std.sort.asc(f64));
-
+            const mid = samples.len / 2;
-        i = 0;
+            const median_ns = if (samples.len % 2 == 0)
-        const increment: @Vector(vector_len, T) = @splat(1);
+                (samples[mid - 1] + samples[mid]) / 2.0
-        while (i < iterations) : (i += 1) {
-            if (comptime @typeInfo(T) == .int)
-                vector_val = vector_val +% increment
             else
-                vector_val = vector_val + increment;
+                samples[mid];
+            return median_ns / @as(f64, @floatFromInt(iters));
         }
-        const vector_time = start_vector.durationTo(getTime()).toMicroseconds();
+    }.f;
 
-        // --- Results ---
+    try writer.print(
-        const s_float = @as(f64, @floatFromInt(scalar_time));
+        \\
-        const v_float = @as(f64, @floatFromInt(vector_time));
+        \\ High Dimension Tensor benchmark — {d} iterations, {d} samples/cell
+        \\ (Results in ns/op)
+        \\
+        \\┌─────────────────┬──────┬──────────────┬──────────────┬──────────────┬──────────────┐
+        \\│ Operation       │ Type │        2x2x2 │        3x3x3 │        4x4x4 │  10x10x10x10 │
+        \\├─────────────────┼──────┼──────────────┼──────────────┼──────────────┼──────────────┤
+        \\
+    , .{ ITERS, SAMPLES });
 
-        // Speedup = ScalarTime / VectorTime.
+    const Types = .{ i32, i64, f32, f64 };
-        // > 1.0 means SIMD is faster.
+    const TNames = .{ "i32", "i64", "f32", "f64" };
-        const speedup = if (vector_time > 0) s_float / v_float else 0;
 
-        try writer.print("│ {d:<10} │ {d:>10} │ {d:>10} │ {d:>9.2}x │\n", .{
+    // Testing multiple structural bounds
-            vector_len,
+    const Shapes = .{
-            scalar_time,
+        &.{ 2, 2, 2 },
-            vector_time,
+        &.{ 3, 3, 3 },
-            speedup,
+        &.{ 4, 4, 4 },
+        &.{ 10, 10, 10, 10 },
+    };
+
+    const Ops = .{ "add", "sub", "mulElem", "mulScalar", "abs" };
+
+    inline for (Ops, 0..) |op_name, o_idx| {
+        inline for (Types, TNames) |T, tname| {
+            try writer.print("│ {s:<15} │ {s:<4} │", .{ op_name, tname });
+
+            inline for (Shapes) |shape| {
+                const V = Tensor(T, .{ .L = 1 }, .{}, shape);
+                const Q = Tensor(T, .{ .T = 1 }, .{}, &.{1}); // For scalar broadcasting operations
+
+                var samples: [SAMPLES]f64 = undefined;
+
+                for (0..SAMPLES) |s_idx| {
+                    const t_start = getTime();
+
+                    for (0..ITERS) |i| {
+                        std.mem.doNotOptimizeAway({
+                            const t1 = V.splat(getVal(T, i, 63));
+
+                            _ = if (comptime std.mem.eql(u8, op_name, "add"))
+                                t1.add(V.splat(getVal(T, i +% 7, 63)))
+                            else if (comptime std.mem.eql(u8, op_name, "sub"))
+                                t1.sub(V.splat(getVal(T, i +% 3, 63)))
+                            else if (comptime std.mem.eql(u8, op_name, "mulElem"))
+                                t1.mul(V.splat(getVal(T, i +% 5, 63)))
+                            else if (comptime std.mem.eql(u8, op_name, "mulScalar"))
+                                t1.mul(Q.splat(getVal(T, i +% 2, 63)))
+                            else if (comptime std.mem.eql(u8, op_name, "abs"))
+                                t1.abs()
+                            else
+                                unreachable;
                         });
-        try writer.flush();
+                    }
 
-        std.mem.doNotOptimizeAway(scalar_val);
+                    const t_end = getTime();
-        std.mem.doNotOptimizeAway(vector_val);
+                    samples[s_idx] = @as(f64, @floatFromInt(t_start.durationTo(t_end).toNanoseconds()));
                 }
-    try writer.print("└────────────┴────────────┴────────────┴────────────┘\n", .{});
+
+                const median_ns_per_op = computeStats(&samples, ITERS);
+                try writer.print(" {d:>12.1} │", .{median_ns_per_op});
+            }
+            try writer.print("\n", .{});
+        }
+
+        if (o_idx < Ops.len - 1) {
+            try writer.print("├─────────────────┼──────┼──────────────┼──────────────┼──────────────┼──────────────┤\n", .{});
+        }
+    }
+    try writer.print("└─────────────────┴──────┴──────────────┴──────────────┴──────────────┴──────────────┘\n", .{});
 }

src/helper.zig

@@ -1,97 +0,0 @@
-const std = @import("std");
-
-pub fn isInt(comptime T: type) bool {
-    return @typeInfo(T) == .int or @typeInfo(T) == .comptime_int;
-}
-
-pub fn printSuperscript(writer: *std.Io.Writer, n: i32) !void {
-    if (n == 0) return;
-    var val = n;
-    if (val < 0) {
-        try writer.writeAll("\u{207B}");
-        val = -val;
-    }
-    var buf: [12]u8 = undefined;
-    const str = std.fmt.bufPrint(&buf, "{d}", .{val}) catch return;
-    for (str) |c| {
-        const s = switch (c) {
-            '0' => "\u{2070}",
-            '1' => "\u{00B9}",
-            '2' => "\u{00B2}",
-            '3' => "\u{00B3}",
-            '4' => "\u{2074}",
-            '5' => "\u{2075}",
-            '6' => "\u{2076}",
-            '7' => "\u{2077}",
-            '8' => "\u{2078}",
-            '9' => "\u{2079}",
-            else => unreachable,
-        };
-        try writer.writeAll(s);
-    }
-}
-
-const Scales = @import("Scales.zig");
-const Dimensions = @import("Dimensions.zig");
-const Dimension = @import("Dimensions.zig").Dimension;
-
-pub fn finerScales(comptime T1: type, comptime T2: type) Scales {
-    const d1: Dimensions = T1.dims;
-    const d2: Dimensions = T2.dims;
-    const s1: Scales = T1.scales;
-    const s2: Scales = T2.scales;
-    comptime var out = Scales.initFill(.none);
-    inline for (std.enums.values(Dimension)) |dim| {
-        const scale1 = comptime s1.get(dim);
-        const scale2 = comptime s2.get(dim);
-        out.set(dim, if (comptime d1.get(dim) == 0 and d2.get(dim) == 0)
-            .none
-        else if (comptime d1.get(dim) == 0)
-            scale2
-        else if (comptime d2.get(dim) == 0)
-            scale1
-        else if (comptime scale1.getFactor() > scale2.getFactor())
-            scale2
-        else
-            scale1);
-    }
-    comptime return out;
-}
-
-// ---------------------------------------------------------------------------
-// RHS normalisation helpers
-// ---------------------------------------------------------------------------
-
-const Quantity = @import("Quantity.zig").Quantity;
-
-/// Returns true if `T` is a `Scalar_` type (has `dims`, `scales`, and `value`).
-pub fn isScalarType(comptime T: type) bool {
-    return @typeInfo(T) == .@"struct" and
-        @hasDecl(T, "ISQUANTITY") and
-        @field(T, "ISQUANTITY");
-}
-
-/// Resolve the Scalar type that `rhs` will be treated as.
-///
-/// Accepted rhs types:
-///   - Any `Scalar_` type               → returned as-is
-///   - `comptime_int` / `comptime_float` → dimensionless `Scalar_(BaseT, {}, {})`
-///   - `BaseT` (the scalar's value type) → dimensionless `Scalar_(BaseT, {}, {})`
-///
-/// Everything else is a compile error, including other int/float types.
-pub fn rhsQuantityType(comptime ValueType: type, N: usize, comptime RhsT: type) type {
-    if (comptime isScalarType(RhsT)) return RhsT;
-    if (comptime RhsT == comptime_int or RhsT == comptime_float or RhsT == ValueType)
-        return Quantity(ValueType, N, .{}, .{});
-    @compileError(
-        "rhs must be a Scalar, " ++ @typeName(ValueType) ++
-            ", comptime_int, or comptime_float; got " ++ @typeName(RhsT),
-    );
-}
-
-/// Convert `rhs` to its normalised Scalar form (see `rhsScalarType`).
-pub inline fn toRhsQuantity(comptime BaseT: type, N: usize, rhs: anytype) rhsQuantityType(BaseT, N, @TypeOf(rhs)) {
-    if (comptime isScalarType(@TypeOf(rhs))) return rhs;
-    const DimLess = Quantity(BaseT, N, .{}, .{});
-    return DimLess{ .data = @splat(@as(BaseT, rhs)) };
-}

src/lib.zig

@@ -0,0 +1,9 @@
+const std = @import("std");
+
+pub const TensorStatic = @import("TensorStatic.zig").Tensor;
+pub const TensorAlloc = @import("TensorAlloc.zig").Tensor;
+pub const TensorGpu = @import("TensorGpu.zig").Tensor;
+pub const Dimensions = @import("Dimensions.zig");
+pub const Scales = @import("Scales.zig");
+pub const Base = @import("Base.zig");
+pub const UnitParser = @import("UnitParser.zig");

src/main.zig

@@ -1,15 +0,0 @@
-const std = @import("std");
-
-pub const Vector = @import("Quantity.zig").Vector;
-pub const Scalar = @import("Quantity.zig").Scalar;
-pub const Dimensions = @import("Dimensions.zig");
-pub const Scales = @import("Scales.zig");
-pub const Base = @import("Base.zig");
-
-test {
-    _ = @import("Quantity.zig");
-    _ = @import("Dimensions.zig");
-    _ = @import("Scales.zig");
-    _ = @import("Base.zig");
-    _ = @import("helper.zig");
-}

src/shared.zig

@@ -0,0 +1,155 @@
+const std = @import("std");
+const Scales = @import("Scales.zig");
+const UnitScale = Scales.UnitScale;
+const Dimensions = @import("Dimensions.zig");
+const Dimension = Dimensions.Dimension;
+
+pub const TensorKind = enum { static, alloc, gpu };
+
+pub fn isTensor(comptime T: type) bool {
+    return comptime @typeInfo(T) == .@"struct" and @hasDecl(T, "ISTENSOR");
+}
+
+pub fn shapeTotal(shape: []const comptime_int) usize {
+    var t: comptime_int = 1;
+    for (shape) |s| t *= s;
+    return t;
+}
+
+/// Check if two shapes are strictly identical.
+pub fn shapeEql(a: []const comptime_int, b: []const comptime_int) bool {
+    if (a.len != b.len) return false;
+    for (a, 0..) |v, i|
+        if (v != b[i]) return false;
+    return true;
+}
+
+/// Row-major (C-order) strides: strides[i] = product(shape[i+1..]).
+///   e.g. shape {3, 4} → strides {4, 1}
+///        shape {2, 3, 4} → strides {12, 4, 1}
+pub fn shapeStrides(shape: []const comptime_int) [shape.len]comptime_int {
+    var st: [shape.len]comptime_int = undefined;
+    if (shape.len == 0) return st;
+    st[shape.len - 1] = 1;
+    if (shape.len > 1) {
+        var i: comptime_int = shape.len - 1;
+        while (i > 0) : (i -= 1) st[i - 1] = st[i] * shape[i];
+    }
+    return st;
+}
+
+/// Return a copy of `shape` with the element at `axis` removed.
+pub fn shapeRemoveAxis(shape: []const comptime_int, axis: comptime_int) [shape.len - 1]comptime_int {
+    var out: [shape.len - 1]comptime_int = undefined;
+    var j: comptime_int = 0;
+    for (shape, 0..) |v, i| {
+        if (i != axis) {
+            out[j] = v;
+            j += 1;
+        }
+    }
+    return out;
+}
+
+/// Concatenate two compile-time slices.
+pub fn shapeCat(a: []const comptime_int, b: []const comptime_int) [a.len + b.len]comptime_int {
+    var out: [a.len + b.len]comptime_int = undefined;
+    for (a, 0..) |v, i| out[i] = v;
+    for (b, 0..) |v, i| out[a.len + i] = v;
+    return out;
+}
+
+/// Decode a flat row-major index into N-D coordinates.
+/// Called only in comptime contexts (all arguments are comptime).
+pub fn decodeFlatCoords(flat: comptime_int, n: comptime_int, strd: [n]comptime_int) [n]usize {
+    var coords: [n]comptime_int = undefined;
+    var tmp = flat;
+    for (0..n) |i| {
+        coords[i] = if (strd[i] == 0) 0 else tmp / strd[i];
+        tmp = if (strd[i] == 0) 0 else tmp % strd[i];
+    }
+    return coords;
+}
+
+/// Encode N-D coordinates into a flat row-major index.
+/// Called only in comptime contexts.
+pub fn encodeFlatCoords(coords: []const usize, n: usize, strd: [n]usize) usize {
+    var flat: usize = 0;
+    for (0..n) |i| flat += coords[i] * strd[i];
+    return flat;
+}
+
+/// Rebuild a full coordinate array by inserting `val` at `axis` into `free`.
+/// `free` holds the remaining (non-contracted) coordinates in order.
+pub fn insertAxis(
+    comptime n: usize,
+    comptime axis: usize,
+    comptime val: usize,
+    comptime free: []const usize,
+) [n]usize {
+    var out: [n]usize = undefined;
+    var fi: usize = 0;
+    for (0..n) |i| {
+        if (i == axis) {
+            out[i] = val;
+        } else {
+            out[i] = free[fi];
+            fi += 1;
+        }
+    }
+    return out;
+}
+
+pub inline fn isInt(comptime T: type) bool {
+    return @typeInfo(T) == .int or @typeInfo(T) == .comptime_int;
+}
+
+pub fn finerScales(comptime T1: type, comptime T2: type) Scales {
+    const d1: Dimensions = T1.dims;
+    const d2: Dimensions = T2.dims;
+    const s1: Scales = T1.scales;
+    const s2: Scales = T2.scales;
+    comptime var out = Scales.initFill(.none);
+    for (std.enums.values(Dimension)) |dim| {
+        const scale1 = comptime s1.get(dim);
+        const scale2 = comptime s2.get(dim);
+        out.set(dim, if (comptime d1.get(dim) == 0 and d2.get(dim) == 0)
+            .none
+        else if (comptime d1.get(dim) == 0)
+            scale2
+        else if (comptime d2.get(dim) == 0)
+            scale1
+        else if (comptime scale1.getFactor() > scale2.getFactor())
+            scale2
+        else
+            scale1);
+    }
+    return out;
+}
+
+pub fn printSuperscript(writer: *std.Io.Writer, n: i32) !void {
+    if (n == 0) return;
+    var val = n;
+    if (val < 0) {
+        try writer.writeAll("\u{207B}");
+        val = -val;
+    }
+    var buf: [12]u8 = undefined;
+    const str = std.fmt.bufPrint(&buf, "{d}", .{val}) catch return;
+    for (str) |c| {
+        const s = switch (c) {
+            '0' => "\u{2070}",
+            '1' => "\u{00B9}",
+            '2' => "\u{00B2}",
+            '3' => "\u{00B3}",
+            '4' => "\u{2074}",
+            '5' => "\u{2075}",
+            '6' => "\u{2076}",
+            '7' => "\u{2077}",
+            '8' => "\u{2078}",
+            '9' => "\u{2079}",
+            else => unreachable,
+        };
+        try writer.writeAll(s);
+    }
+}

src/test.zig

@@ -0,0 +1,9 @@
+test {
+    _ = @import("TensorStatic.zig");
+    _ = @import("TensorAlloc.zig");
+    _ = @import("TensorGpu.zig");
+    _ = @import("Dimensions.zig");
+    _ = @import("Scales.zig");
+    _ = @import("Base.zig");
+    _ = @import("UnitParser.zig");
+}