Products

What We Build

We build bold, open-source technology for the AI and HPC era, from a full datacenter you can spin up in software, to the training that grows the engineers who run it, to sensing that guards the ground it stands on. Dreamed up and engineered in Ireland, shared openly with the world.

Cnuas Virtual AI HPC Infrastructure

A complete AI/HPC datacenter, in software. Cnuas is a 100% software-emulated infrastructure built on QEMU, with virtual GPUs, virtual NVLink-equivalent fabric, virtual RoCE and InfiniBand TOR switches and PCIe-attached virtual NICs, running unmodified Linux software stacks inside guests.

Anyone with a laptop can stand up a multi-node GPU cluster, exercise the same libibverbs and CUDA-equivalent APIs used in production, and learn or prototype against a real-feeling fabric, without owning a single InfiniBand cable.

  • HiGPU, virtual GPU PCIe device with SIMT execution model
  • HiLink, GPU-to-GPU peer fabric (NVLink-equivalent)
  • HiSwitch, virtual TOR switch, RoCEv2 and InfiniBand multiplexed
  • HiNIC, virtual RoCE-IB NIC kernel driver and verbs provider
  • cnuas, unified CLI and REST API to configure and drive the whole system
  • HCC, CUDA-equivalent compute stack with the hcc device compiler and HiIR
  • HiCCL and HiSHMEM, NCCL and NVSHMEM equivalents over the HiLink fabric
  • hi-smi, hi-dcgmi and hi-prof, nvidia-smi, DCGM and Nsight-equivalent management tools
  • Math libraries, HiBLAS, HiDNN, HiFFT, HiSPARSE and HiSOLVER (cuBLAS and cuDNN equivalents)
  • Standard Linux Verbs API, so guests use libibverbs unmodified
  • Modelled on Open Compute Project Open Rack v3, the hyperscaler reference design
Explore Cnuas Documentation GitHub Repo ↗ White Paper (PDF)
🖥️

HiGPU with HiLink

SIMT virtual GPU and the inter-GPU fabric that connects them. The full NVIDIA-equivalent compute stack, with a CUDA Runtime equivalent (HCC), an nvidia-smi equivalent (hi-smi), and an LLVM-backed toolchain (hcc).

🔀

HiSwitch with HiNIC

Top-of-rack switch daemon and PCIe-attached virtual NIC. Handles RoCEv2 over UDP and full InfiniBand including LRH routing, with DCB (PFC, ETS and ECN) modelled end to end.

📚

Used by HiCAIN

Cnuas is the lab environment for HiCAIN courses, where students run real distributed-training workloads against the emulated fabric, no hardware required.

HiCAIN
Academic Courses & Training Programme
Education CUDA ROCm RDMA OFED MPI

HiCAIN

High-performance Computing, AI and Networking, PacketFive's open-source training programme for the full AI/HPC software stack. Delivered through the HiCAIN Academy learning platform, HiCAIN gives students, engineers and practitioners the deep technical skills modern AI clusters demand.

  • Structured courses on CUDA, ROCm, RDMA, RoCE, OFED, Open MPI and InfiniBand
  • HiCAIN Academy, a full LMS with progress tracking and certificates of completion
  • Practical labs designed around the Cnuas Virtual AI HPC Infrastructure (no expensive hardware needed)
  • Instructor and self-paced learning tracks
  • Google SSO, JWT auth, and email verification built-in
  • Open-source content and platform under permissive licences
HiCAIN.io ↗ HiCAIN Academy ↗

TORUS

Passive, layered sensing for AI/HPC data center facilities. TORUS is an Unattended Facility Sensing platform for hyperscale, colocation and enterprise AI/HPC data centers. It monitors campus perimeters, data halls and critical approaches across ground, visual and drone/UAS detection layers, classifies security events with AI at the edge and at command, and drives facility response under safety interlock, even where mains power and network are unavailable.

  • TORUS-SN, the ground sensor node, with seismic and acoustic sensing, on-device ML, onboard GPS and MEMS sensors for node tracking, and optimized battery management
  • Node variants, four classes on one event contract (fixed wireless, SDR, wired optical, transducer)
  • TORUS-MEG, cluster-aware field gateway with NVIDIA Jetson Orin edge compute and flexible backhaul
  • TORUS-CCISRT, the command layer, one live common operating picture with a safety-interlocked response
  • AI/ML engine, intelligence at three tiers, with TinyML on the node, Jetson at the gateway and correlation at command
  • TORUS Isaac Sim, a cyber-physical deployment simulator to rehearse scenarios before emplacement
Try it! ↗ Capability Brief ↗ Brochure (PDF) GitHub Repo ↗

Datasheets & brochures

TORUS
Unattended Facility Sensing
Ground Visual Drone/UAS Edge AI TIA-942 / OCP Off-grid
🛰️

Layered & passive

Detection across three physical domains, correlated into a classified track with location and confidence, sensing that emits nothing to find.

⚡

Off-grid & unattended

Battery and solar operation, long-range radio, months between visits, auditable by design, with every detection and command timestamped and logged.

Open Source Commitment

All PacketFive R&D output, including code, training material, papers and reference implementations, is published on GitHub under permissive open-source licences.

PacketFive on GitHub ↗