dspk-ops 0.0.1

Signal-processing operations for dspk, written once in CubeCL: filters, resampling, linear algebra, statistics, peak finding, spatial operators
Documentation

dspk

[!WARNING] Early development (0.0.x). APIs change between releases. Documentation is being rewritten.

Multi-channel signal processing written once in CubeCL and run on any GPU (wgpu: Vulkan, Metal, DirectX 12; CUDA; ROCm / HIP) or the CPU, with the device chosen at run time. Processing stages keep data on the device between them, so a recording crosses disk → RAM → device memory as few times as possible.

dspk-app (desktop workbench, not on crates.io): a Neuropixels recording (385 channels, 30 kHz) drawn from min/max pyramids built by dspk-view.

Crates

Crate Contents
dspk Entry crate: re-exports the crates below; features choose what is included.
dspk-core Recordings (RecordingSource), buffers, exact time and sample rates, windows, compute targets.
dspk-io Raw binary and Zarr recordings, format detection, chunk cache, .npy/.npz/Zarr containers.
dspk-ops IIR / FIR / non-linear filters, resampling, spatial operators, linear algebra, statistics, peak finding, device pipelines.
dspk-view Min/max envelopes and pyramids for viewing, local or remote.
dspk-stream Network sessions for continuous signals (QUIC + TLS, protobuf).

Electrophysiology (spike detection, sorters, neural file formats) lives in the synapse/ crates of the repository and is not published yet.

Runtimes

Runtimes are crate features; the device is picked at run time with dspk::core::ComputeTarget (or the DSPK_RUNTIME environment variable) among those compiled in.

[dependencies]
dspk = "0.0.1"                                               # wgpu (default)
# dspk = { version = "0.0.1", features = ["cuda", "cpu"] }   # add CUDA and the CPU runtime
Feature Runtime Tested
wgpu (default) WebGPU: Vulkan, Metal, DirectX 12 NVIDIA RTX 2070 Max-Q, Intel UHD 630 (Vulkan)
cuda NVIDIA CUDA NVIDIA RTX 2070 Max-Q
cpu CubeCL CPU runtime (MLIR / LLVM) x86-64
hip AMD ROCm / HIP compiles; untested (no AMD GPU available)

CUDA and HIP are loaded at run time, so building with them needs no toolkit installed.

Device size limit. Kernels index with 32-bit integers (WebGPU has no 64-bit integers, and one kernel source serves every runtime): one window or batch on the device holds at most 2³² − 1 values (e.g. 384 channels × 11.1 million samples). Larger requests are refused with an error.

License

MIT OR Apache-2.0, at your option (LICENSE-MIT, LICENSE-APACHE).