Type System

Eä is statically typed with no implicit conversions. Every variable, parameter, and expression has a concrete type known at compile time.

Scalar Types

Integer literals default to i32. Float literals default to f32. Use explicit casts (to_f64(x), to_i64(x)) to convert between types.

Vector Types

Vector types hold multiple lanes of the same scalar type. Element-wise operations use dot-operators (.+, .-, .*, ./, etc.).

128-bit Vectors -- SSE (x86) / NEON (ARM)

256-bit Vectors -- AVX2 (x86 only)

These types produce a compile error on ARM targets.

512-bit Vectors -- AVX-512 (x86, --avx512 flag required)

f32x16, f64x8, i32x16, and u64x8 lower to AVX-512F instructions. i8x64, u8x64, i16x32, and u16x32 additionally require AVX-512BW (byte/word SIMD), which is present on Skylake-SP, Ice Lake, Zen 4, and later. Eä emits both feature flags when --avx512 is set.

Using these types without --avx512 produces a compile error.

Half-precision and sub-128-bit narrow widths

The lexer also accepts f16x4 / f16x8 (half-precision float vectors) and sub-128-bit narrow widths used by ARM NEON widening intrinsics: i8x4, i8x8, u8x8, i16x4, u16x4, i32x2. These have target-specific constraints (f16 requires --fp16 on aarch64 and is unavailable on plain x86; the narrow widths exist primarily as inputs to widening multiplies like wmul_i32(i16x4, i16x4) -> i32x4) that are documented alongside the intrinsics that consume them rather than as standalone table rows. See ARM / NEON reference for f16 and All Intrinsics for the narrow-width consumers.

Pointer Types

Pointers represent caller-provided memory. Eä never allocates -- all memory comes from the host language.

The restrict qualifier tells the compiler that the pointer does not alias other pointers, enabling stronger optimizations. Use it when you can guarantee non-overlapping memory.

Struct Types

User-defined value types declared with struct:

struct Pixel {
    r: f32,
    g: f32,
    b: f32,
    a: f32,
}

Structs are passed by value. Access fields with dot syntax: pixel.r. Structs can contain scalar types, vector types, and other structs.

Type Rules

• No implicit conversions between types. Use to_f32(), to_i32(), etc.
• No generics or polymorphism. Write separate functions for each type.
• Vector dot-operators require both operands to have the same vector type.
• Comparisons on vectors (.==, .<, .>) produce boolean vectors, not scalar bools.