Kernels
Hand-tuned GPU kernels for the operations AI spends its time on, checked against high-precision references before they ship.
Riptide builds the compute layer under AI: GPU kernels, a compiler and a runtime that get more out of every chip in your fleet.
Most AI runs on general-purpose kernels that leave performance on the table. We write the code that runs on the chip, and the tools that put it there.
Hand-tuned GPU kernels for the operations AI spends its time on, checked against high-precision references before they ship.
Turns models into code shaped for each chip: tiling, memory layout and scheduling chosen for the hardware, not a lowest common denominator.
Launches work with as little overhead as possible, so small models and decode steps stop waiting on the host.
FlashAttention-style forward and backward passes, full and causal, on the GPU's matrix cores. Plugs into PyTorch's standard attention call.
Step replay that cuts per-launch cost, and fused decode for small models, where reading the weights takes a fraction of the time the launches do today.
Softmax, normalization and elementwise chains fused into single passes, so they run closer to the memory bandwidth the chip actually has.
The core of every model, next on the list.
riptide-amdOpen-source GPU kernels for AMD Instinct, Apache-2.0. Install it next to PyTorch for ROCm and opt in. No fork, no patches, no changes to your stack.
# coming soon pip install riptide-amd import riptide_amd riptide_amd.enable() # attention now runs on riptide
Speed-up versus AMD's best kernel for each case: the faster of PyTorch's FLASH_ATTENTION (AOTriton) and EFFICIENT_ATTENTION backends, measured in the same session. Median of 30 runs, 95% CI, MI350X, bf16, 32 heads, ROCm 10.2, PyTorch 2.15, 2026-11-19.
We're also building kernels and ports for other accelerators and chip makers. That work happens privately, across GPU and accelerator vendors, and we talk about it when it's ready to share.
Kernels and ports for your accelerator, measured honestly against the best available baseline.
More throughput and less energy from the GPUs you already run, without changing your stack.