doca-gpi
Skill for DOCA GPI programming so CUDA kernels drive RDMA queues.
About
Use this skill for hands-on DOCA GPI programming — wiring a GPU-Packet-Initiator context so a CUDA kernel drives RDMA queues directly from GPU memory without host CPU mediation. Covers picking GPI vs doca-gpunetio, the doca_gpi / domain / channel object model, the GPU-side handle handoff (doca_gpu_gpi_channel*), attaching GPU memory to a GPI domain, the domain and channel attribute objects, and debugging DOCA_ERROR_* from doca_gpi_* calls. Trigger even when the user does not explicitly mention "DOCA GPI" — implicit phrasings include "my CUDA kernel needs to post RDMA directly from GPU memory", "DOCA_ERROR_* from doca_gpi_gpu_channel_get", "how do I hand a GPU handle to my CUDA kernel", "how many channels can a GPI domain hold", or "GPU kernel driving RDMA without the host CPU on the path". Refuse and route elsewhere for the doca-gpunetio Send/Receive surface, the doca-rdma queue lifecycle, DPA-side initiation (doca-rdmi), or the CUDA programming model — those belong to other skills.
At a glance
- Made for
- Developers
- Pricing
- Free
- Access
- No sign-in
Category, summary and these details are StealthStack's reading of the publisher's own listing. Check the publisher's page before you rely on them.
Capabilities
The crawler did not record capability metadata for this resource. Inspect the endpoint directly to see what it exposes.
Provenance
- Discovered
- Relayed by agntcy
- Identifier
- urn:air:outshift.io:agntcy:doca-gpi
- Catalog host
- outshift.io · via agntcy registry
- Anchor check
- Not anchored
- Last crawled
- seen 1h ago
Discovered through outshift.io's registry, not anchored by StealthStack. We relay the listing as-is; we have not checked that the URN authority matches the publishing host. How trust works →