- XENON Workstations — GPU-optimised towers and desktop AI supercomputers, configured to your applications and budget.
- XENON Servers — Custom rack servers on Intel, AMD and NVIDIA Grace platforms, up to ten GPUs per node.
- XENON Storage — Dense, efficient rackmount storage and NAS, sized to your data growth.
AI Factories
AI factories are built from GPU-dense nodes, and XENON Customs builds them to order: NVIDIA HGX 4- and 8-GPU systems, servers carrying up to ten PCIe GPUs, and NVIDIA Grace and Grace Hopper platforms. XENON assembles them into production AI clusters with the XENON Cluster Stack, ready for training, fine-tuning and inference.
HPC
Custom single- and dual-socket Intel Xeon and AMD EPYC nodes, with memory to 8TB and InfiniBand or high-speed Ethernet fabrics, form the building blocks of XENON-built research clusters. High-density blade systems pack up to 120 nodes into a single rack. Every configuration is matched to your codes: CFD, molecular dynamics, Cryo-EM, rendering or simulation.
Data Management
XENON Customs storage servers scale to 36 drive bays in a chassis, from all-flash NVMe nodes for software-defined storage to high-capacity NAS appliances with redundant power and cooling. They slot into XENON-designed data management architectures alongside our partner platforms, giving your data the tier it deserves at each stage of its life.
Services
Every XENON Customs system starts with an engineer, not a catalogue: we scope your applications, budget and constraints, then design, build, burn-in and test in Melbourne. Systems ship with your OS and software stack installed, including CUDA and AI frameworks. After delivery, you talk directly to the team that built your system, with return-to-base, next-business-day onsite and 24×7 global support options.