| tpt-simd-scatter cargo | 0.1.1 | none | none | SIMD non-contiguous stores (scatter) | 23 | 3d ago |
| tpt-simd-saturate cargo | 0.1.1 | none | none | SIMD saturating arithmetic and saturating pack operations | 40 | 3d ago |
| tpt-simd-rounding cargo | 0.1.1 | none | none | SIMD rounding operations | 27 | 3d ago |
| tpt-simd-rng cargo | 0.1.1 | none | none | Lane-parallel xoshiro256++ and Philox4x32-10 generators with SIMD-friendly uniform and normal (Box-Muller) slice fills | 26 | 3d ago |
| tpt-simd-pixel cargo | 0.1.1 | none | none | Pixel kernels: RGB/RGBA conversion, planar/interleaved layout, u8 and f32 conversion, premultiply, HSV, histograms, with no runtime dependencies | 23 | 3d ago |
| tpt-simd-ntt cargo | 0.1.1 | none | none | Montgomery and Barrett modular arithmetic and an in-place number-theoretic transform for 31-bit primes | 23 | 3d ago |
| tpt-simd-mul cargo | 0.1.1 | none | none | SIMD multiply variants (high/low/widening/complex) | 33 | 3d ago |
| tpt-simd-math cargo | 0.1.1 | none | none | Vectorised f32 exp, ln, sin, cos, tanh and erf: branch-free polynomial approximations with documented ULP error | 27 | 3d ago |
| tpt-simd-interpolate cargo | 0.1.1 | none | none | SIMD interpolation kernels | 26 | 3d ago |
| tpt-simd-horizontal cargo | 0.1.1 | none | none | SIMD horizontal reductions (sum, min, max, product) | 61 | 3d ago |
| tpt-simd-hash cargo | 0.1.1 | none | none | Checksums and hashes: CRC-32, Adler-32, FNV-1a and a rolling hash, with no runtime dependencies | 26 | 3d ago |
| tpt-simd-half cargo | 0.1.1 | none | none | f16/bf16 conversions (wrapping the half crate), slice conversions, and int8 / Q8_0 block quantisation | 23 | 3d ago |
| tpt-simd-geom cargo | 0.1.1 | none | none | Batched 3D geometry over structure-of-arrays f32 data: distance, dot, AABB overlap, ray-AABB | 23 | 3d ago |
| tpt-simd-gather cargo | 0.1.1 | none | none | SIMD non-contiguous loads (gather) | 36 | 3d ago |
| tpt-simd-fixed cargo | 0.1.1 | none | none | Type-safe SIMD fixed-point arithmetic | 25 | 3d ago |
| tpt-simd-distance cargo | 0.1.1 | none | none | Vector distances and similarities: inner product, squared L2, cosine, Hamming | 23 | 3d ago |
| tpt-simd-dct cargo | 0.1.1 | none | none | Table-based DCT-II/III and MDCT/IMDCT for codec-sized blocks | 23 | 3d ago |
| tpt-simd-core cargo | 0.1.1 | none | none | Core traits, type aliases and feature detection for tpt-simd | 184 | 3d ago |
| tpt-simd-convolve cargo | 0.1.1 | none | none | SIMD convolution and FIR primitives | 25 | 3d ago |
| tpt-simd-compare cargo | 0.1.1 | none | none | SIMD comparisons producing masks | 25 | 3d ago |
| tpt-simd-butterfly cargo | 0.1.1 | none | none | SIMD FFT/DCT butterfly operations | 25 | 3d ago |
| tpt-simd-blend cargo | 0.1.1 | none | none | SIMD branch-free conditional selection | 34 | 3d ago |
| tpt-simd-blas cargo | 0.1.1 | none | none | BLAS-style f32/f64 kernels: level-1, gemv and a packed, blocked gemm | 96 | 3d ago |
| tpt-simd-bits cargo | 0.1.1 | none | none | Bit-level and integer-stream kernels: popcount, bitset ops, zigzag, delta and prefix sum | 23 | 3d ago |
| tpt-simd-audio cargo | 0.1.1 | none | none | Audio kernels: mix, interleave, peak/RMS, FIR, biquad, and windowed-sinc resampling, with no runtime dependencies beyond libm | 23 | 3d ago |
| tpt-simd-aligned cargo | 0.1.1 | none | none | SIMD-aligned memory types and helpers | 25 | 3d ago |
| agent-ext cargo | 0.1.0 | none | none | Agent extension points: SKILL.md discovery with a deferred skill tool, and Claude-Code-style shell hooks (JSON on stdin, exit 2 blocks) that fail open | 9 | 3d ago |
| agent-memgc cargo | 1.0.2 | none | none | High-performance linter, cache-layout optimizer, and generational mark-and-sweep garbage collector for AI agent instruction files (AGENTS.md, CLAUDE.md,… | 28 | 3d ago |
| airframe cargo | 0.4.4 | none | none | Pure-Rust WebGPU inference engine for GGUF models. No C++ toolchain, no llama.cpp. Runs on NVIDIA, AMD, Intel, and Apple Silicon via wgpu. | 3.1K | 3d ago |
| newt-skills cargo | 0.8.2 | none | none | Newt-Agent agentskills.io-compatible skills — discovery + SKILL.md parsing for progressive disclosure | 549 | 3d ago |
