Products
Compute for every layer of the rack.
Every product is quoted against confirmed supply as one landed AUD price, with freight, clearance and GST included and lead time stated.
GPU servers
8-GPU HGX systems for training and large-model inference.
NVIDIA HGX H200 8-GPU server
8 x H200 SXM, 1,128 GB of HBM3e in one node.
- GPUs
- 8 x NVIDIA H200 SXM
- GPU memory
- 141 GB HBM3e per GPU (1,128 GB total)
- Memory bandwidth
- 4.8 TB/s per GPU
NVIDIA HGX H100 8-GPU server
8 x H100 SXM, the proven training workhorse.
- GPUs
- 8 x NVIDIA H100 SXM
- GPU memory
- 80 GB HBM3 per GPU (640 GB total)
- Memory bandwidth
- 3.35 TB/s per GPU
NVIDIA HGX B300 8-GPU server
Blackwell Ultra, 288 GB of HBM3e per GPU.
- GPUs
- 8 x NVIDIA B300 (Blackwell Ultra)
- GPU memory
- 288 GB HBM3e per GPU (about 2.3 TB total)
- GPU interconnect
- NVLink 5, 1.8 TB/s per GPU
GPUs
Workstation and data-centre PCIe GPUs for inference, rendering and research.
NVIDIA RTX PRO 6000 Blackwell
96 GB of GDDR7 in a single PCIe card.
- GPU memory
- 96 GB GDDR7 with ECC
- Memory bandwidth
- 1,792 GB/s
- CUDA cores
- 24,064
NVIDIA H200 NVL
H200 memory in a PCIe card for standard servers.
- GPU memory
- 141 GB HBM3e
- Memory bandwidth
- 4.8 TB/s
- Interface
- PCIe 5.0 x16, dual-slot
NVIDIA GeForce RTX 5090
32 GB GDDR7, the fastest consumer GPU.
- GPU memory
- 32 GB GDDR7
- Memory bandwidth
- 1,792 GB/s
- CUDA cores
- 21,760
Server memory
DDR5 registered ECC memory, matched to your platform.
Enterprise storage
Data-centre NVMe in U.2, E1.S and E3.S form factors.
Data-centre infrastructure
Liquid cooling, optics and power for AI racks.