01 · GPU
GPUs for training, inference and visualisation.
Data-centre accelerators, professional workstation cards and inference-optimised boards.
PCIe · SXM24–192 GB HBM / GDDRTraining · Inference · Viz
Data-centre accelerators
Passive-cooled boards for rack servers.
| Form factor | PCIe dual-slot / SXM module |
|---|---|
| Memory | 40–192 GB HBM |
| Interconnect | PCIe Gen5, NVLink-class bridges |
| Power | 300–700 W |
| Cooling | Passive, server airflow |
Workstation / professional
Active-cooled cards for CAD, rendering and local ML development.
| Form factor | PCIe dual-slot, active |
|---|---|
| Memory | 16–96 GB GDDR/ECC |
| Displays | 4× DP |
| Power | 150–300 W |
Inference cards
Low-profile and single-slot boards for edge and density deployments.
| Form factor | Single-slot / low-profile |
|---|---|
| Memory | 16–48 GB |
| Power | 70–150 W |
| Use | Inference, transcoding, VDI |
Use cases
LLM training clustersInference servingRendering farmsCAD/CAE workstationsVDI
Request for quotation
Send the requirement. We return availability, lead time and price.
Single units to full deployments. Compatibility and warranty reviewed before quotation.