32 posts
OxiONNX 0.1.7 grows oxionnx-cuda from 25 to 40 accelerated ops with real Conv support, session-lifetime weight/buffer residency, and batched MatMul up to 17x faster — plus a fix for a two-stream race that silently returned wrong GEMM results on real hardware. 3,595 tests passing. The sovereign Pure Rust ONNX layer for COOLJAPAN.
ToRSh 0.2.0 completes autograd backward coverage, consolidates GPU compute onto the pure-Rust oxicuda stack, ships full NumPy/pandas/SciPy interop and PyTorch-compatible LR schedulers, and replaces fabricated success paths with honest errors. 10,638 tests passing. The sovereign deep-learning layer for COOLJAPAN.
OxiCUDA 0.5.5 audits oxicuda-metal and oxicuda-webgpu on real Apple Silicon, fixing conv2d/attention/softmax false completions where finished MSL kernels sat unused behind CPU scalar loops. Separately, a downstream investigation wires up an unreachable split-K GEMM path, replaces a silent Winograd no-op with a real kernel, and fixes wrong Ampere GA10x hardware constants — measured on a real RTX A4000. 38,987 tests, ~1.31M SLoC, 74 crates.
OxiCUDA 0.5.4 adds WarpVec/WarpMask, a SIMD-flavored warp-vector layer for oxicuda-ptx's builder DSL, six new IR instructions for warp shuffle/vote, and a foreign-compiler PTX interop test proving the driver/launch stack hosts upstream rustc-generated PTX too. 38,689 tests, ~1.30M SLoC, 74 crates.
OxiCUDA 0.5.2 fixes a class of PTX portability bugs that CUDA 11.x silently tolerated and CUDA 12.9+ toolchains reject outright (non-ASCII bytes in generated comments), closes several Windows-specific test and lock-contention bugs, and fixes a numerical-correctness bug in the tensor-network DMRG excited-states solver. 38,675 tests, ~1.30M SLoC, 74 crates.
OxiCUDA 0.5.3 fixes an async-copy race condition where cuMemcpyDtoHAsync/cuMemcpyHtoDAsync and DeviceBuffer::copy_from_host could return before the transfer had actually landed, letting the very next read observe stale or zeroed device/host memory. 38,675 tests, ~1.30M SLoC, 74 crates.
OxiFFT 0.4.1 is a small, focused release: it pulls in OxiCUDA 0.5.3, which fixes a race condition where async CUDA memory copies in oxicuda-fft could return before the transfer landed — risking silently corrupted or all-zero GPU transform results under load.
OxiBonsai 0.2.3 ships a two-pass, 109-agent production-hardening audit: a P0 fix for a silent CUDA prefill/decode KV-cache desync, full Metal GPU coverage for Q4_0/Q8_0/K-quant and FP8 batch prefill, real OpenAI-compatible penalties and logprobs, and tokenizer fidelity fixes for CJK/emoji text — 91 confirmed findings closed, 5,158 tests passing. Sub-2-bit Pure Rust sovereign AI inference for the COOLJAPAN ecosystem.
OxiCUDA 0.5.1 adds oxicuda-nvrtc, a pure-Rust runtime loader for NVIDIA's NVRTC CUDA-C-to-PTX JIT compiler, completing the zero-SDK-dependency story alongside oxicuda-driver. Graceful degradation, process-wide caching, and direct PTX handoff to oxicuda-driver. 38,646 tests, ~1.30M SLoC, 74 crates.
OxiCUDA 0.5.0, the sovereign GPU-compute layer for the COOLJAPAN ecosystem, fixes F64-precision PTX codegen bugs that made ptxas reject elementwise and reduction kernels, closes a mixed-precision GEMM accumulator bug, and hits zero unwrap()/expect() in library code. 38,622 tests, ~1.30M SLoC, 73 crates.
sklears 0.2.0 ships sklears-core::gpu, a real oxicuda-backed CUDA foundation (GpuBackend, GpuArray, GpuMatrixOps) powering on-device GEMM, Cholesky/LU/QR/SVD solves, and HNSW k-NN across 9 downstream crates, plus a wide correctness sweep. 12,721 tests passing, 36 crates, >99% scikit-learn API coverage held.
TrustformeRS 0.2.0 gives CUDA the same device-resident attention pipeline Metal already had, adds batched CUDA matmul and refcounted GPU buffers, fixes a silent CUDA data race, and drops the torch/libtorch backend workspace-wide. The sovereign transformer layer for the COOLJAPAN ecosystem.
OxiCUDA 0.4.1 extends on-device GPU validation into BLAS, DNN, and the driver/memory/launch stack, catching concurrency races that single-threaded testing can't. Plus the first correctness pass across five non-CUDA backends and real driver-backed CUDA Graph capture. 38,612 tests, ~1.28M SLoC, 73 crates.
TrustformeRS 0.1.4 migrates CUDA and Metal from cudarc/scirs2-MPS to the Pure-Rust oxicuda stack, passes 12/12 CPU↔CUDA parity tests on a real RTX A4000, ships real PyRwkvModel/PyMambaModel classes on PyO3 0.28, and goes unwrap()-free workspace-wide. The sovereign transformer layer for the COOLJAPAN ecosystem.
OxiCUDA 0.4.0 is an on-device validation pass: for the first time, hand-written PTX across 60+ crates was JIT-compiled and run on real NVIDIA hardware (RTX A4000, sm_86, CUDA 12.4) instead of only checked for CPU-logic parity — catching register-shadowing bugs, base-2/base-e math errors, invalid PTX, and literal stub kernels. 38,093 tests passing, ~1.27M SLoC, 73 crates.
SciRS2 0.6.0 introduces the pure-Rust oxicuda-* CUDA stack as a direct, per-crate NVIDIA performance backend and decentralizes GPU out of scirs2-core. Ten crates — fft, symbolic, interpolate, special, stats, graph, linalg, optimize, datasets, and vision — gain an off-by-default, runtime-probed, f64-native cuda feature standing alongside the existing wgpu/WebGPU portability path, now standardized under one wgpu feature name across the ecosystem. A default build still compiles zero oxicuda. Pure Rust, Apache-2.0.
ToRSh 0.1.3 lands the oxicuda GPU backend (no CUDA SDK at build time), eliminates the last C/asm dependency, ships a bandwidth-optimal ring all-reduce, completes the Node.js N-API layer, and delivers 15–30% throughput gains from phase-4 chunking helpers.
OxiCUDA 0.2.0, the pure-Rust replacement for the NVIDIA CUDA Toolkit, lands the 'Wave AAA+64' expansion: adaptive RK45 with Richardson extrapolation, Extended Persistence and Discrete Morse theory (oxicuda-tda), Parametric UMAP, and Fisher Information estimation — plus a workspace-wide zero-unwrap reliability pass and 32,320 passing tests. ~783K lines across 73 crates. No CUDA SDK, no nvcc.
Pure-Rust legal statute engine. 0.1.6 adds real NVIDIA CUDA GPU acceleration for population-scale simulation (optional, with transparent CPU fallback), a hardened security/governance API layer, an autonomous and post-quantum-safe compliance audit subsystem, legal analytics with risk heatmaps, French civil/company law, and a fully C-free storage backend via OxiSQL. 18,398 tests passing.
OxiBonsai 0.2.2 adds `oxibonsai repl`: a resident ImageSession that loads the DiT, VAE, and text encoder once and iterates on prompts without re-paying the load/dequant cost — with images shown inline in Ghostty via a pure-Rust kitty graphics protocol, a `:fast`/`:hq` preview→finalize loop, and documented per-platform GPU flags. Sub-2-bit Pure Rust sovereign AI inference for the COOLJAPAN ecosystem.
OxiBonsai 0.2.0 opens the 0.2 series: a concurrent engine pool that shares one 1.16 GB embedding table across replicas, a CPU↔Metal byte-identical parity guard, a parity-first CUDA imagen backend (~3.2× to ~31.7s on A4000), --seed byte-exact reproducible images, and a stable-toolchain build — sub-2-bit Pure Rust sovereign AI inference for the COOLJAPAN ecosystem.
Pure-Rust CUDA Toolkit replacement: a maintenance release with numerical-stability refinements in the HMC variational sampler, stream-ordered allocator tuning, and TriMap reduction polish — 23,535 passing tests. No CUDA SDK, no nvcc.
Pure-Rust replacement for the entire NVIDIA CUDA Toolkit. 0.1.7 adds a SYR2K Tensor Core kernel (fused A×Bᵀ + B×Aᵀ rank-2k update) to oxicuda-blas, cross-subsystem CUDA kernel enhancements, and Multi-Operation Scheduling improvements. No CUDA SDK, no nvcc, no C/C++ toolchain.
Pure-Rust replacement for the NVIDIA CUDA Toolkit. OxiCUDA 0.1.6 adds a Tensor Core fast path for SYRK in oxicuda-blas and sixteen new ML crates (adversarial, SSL, continual, multimodal, 3D geometry, PINN, ANN, anomaly, causal, meta, MoE, NeRF, quantum, recsys, RLHF, tabular). No CUDA SDK, no nvcc.
OxiBonsai 0.1.3 makes sub-2-bit serving smarter: a prefix-cache-aware engine that reuses KV-cache across requests with byte-identical cold/warm parity, runtime tokenizer auto-detection, and a GPU weight cache that uploads once. Sub-2-bit Pure Rust sovereign AI inference for the COOLJAPAN ecosystem.
The pure-Rust NVIDIA CUDA Toolkit replacement adds nine new GPU deep-learning crates — generative diffusion, graph neural nets, Mamba SSMs, vision transformers, audio/speech, time-series, Bayesian DL, federated learning, and NAS — growing to ~320K lines across 37 crates with 9,568 passing tests. No CUDA SDK, no nvcc.
Five days after its 1-bit debut, OxiBonsai grows GPUs: a native CUDA NVRTC backend (~21.9 tok/s on Ternary-Bonsai-1.7B, RTX 3060) and a fused Metal full-forward path (~50 tok/s, ~13x speedup) — plus the new ternary TQ2_0_g128 quant family, with NEON/AVX2/AVX-512 GEMV so it flies on CPU too. Sub-2-bit Pure Rust sovereign AI inference for the COOLJAPAN ecosystem, still with no llama.cpp, no BLAS, no C/Fortran.
A small maintenance release for OxiCUDA, the pure-Rust replacement for the NVIDIA CUDA Toolkit. Workspace-wide documentation and quality improvements, with all 28 crates aligned to 0.1.4 so the stack ships in lockstep. The only runtime dependency is the NVIDIA driver.
A quality-and-docs maintenance release for the pure-Rust NVIDIA CUDA Toolkit replacement — workspace-wide polish, internal version alignment to 0.1.3, and continued growth to ~260K lines of safe Rust across 28 crates. The only runtime dependency is still the NVIDIA driver.
Complete, type-safe, memory-safe rewrite of the entire NVIDIA CUDA Toolkit in pure Rust. cuBLAS/cuDNN/cuFFT/cuSPARSE/cuSOLVER/cuRAND and more — all in 253k SLoC across 28 crates. Only runtime dependency is the NVIDIA driver. PTX codegen + autotuner, 7 GPU backends (Metal/Vulkan/WebGPU/ROCm/LevelZero). ≥90–95% of native CUDA performance. The sovereign GPU computing layer for SciRS2 and the entire COOLJAPAN ecosystem (now 21M+ SLoC total).
First patch on the pure-Rust NVIDIA CUDA Toolkit replacement: six new oxicuda-blas elementwise activations (HardSigmoid, HardSwish, Softplus, LeakyRelu, Ceil, Floor) plus substantial ROCm/Vulkan/WebGPU backend growth. ~248K lines across 28 crates.
OxiCUDA 0.1.0 is a pure-Rust, type-safe, memory-safe replacement for the entire NVIDIA CUDA Toolkit software stack — cuBLAS, cuDNN, cuFFT, cuSPARSE, cuSOLVER, cuRAND and more in ~239K lines across 28 crates. The only runtime dependency is the NVIDIA driver. PTX code generation plus a built-in autotuner, all from safe Rust.