Australian Authorised Reseller

Enterprise-Grade
AI Hardware,
Delivered.

Tensor Works sources and supplies the compute infrastructure that powers serious AI workloads — from single GPU nodes to full DGX clusters — for government, research and enterprise buyers across Australia.

Browse Hardware Speak to an Expert
Dicker Data Partner 100% Australian Owned Fast Fulfilment
NVIDIA H100 SXM5
80 GB HBM3 · 3.35 TB/s · NVLink 4
In Stock
NVIDIA DGX H200
8× H200 · 1.1 TB HBM3e · 640 GB total
On Order
InfiniBand NDR400
400 Gb/s · Sub-µs latency · Scale-out fabric
In Stock
AI Enterprise Suite
Full-stack licences · Support · Deployment
Available
NVIDIA Authorised
Dicker Data Partner
Australian Owned
100% Australian Owned
Dedicated Account Manager
Hardware Catalogue

Enterprise AI Compute

Purpose-built hardware for training, inference, simulation and HPC workloads at any scale.

GPU Accelerator
NVIDIA H100
The world's most powerful AI training GPU — transformer engine, NVLink 4, PCIe and SXM5 form factors.
Memory80 GB HBM3
Bandwidth3.35 TB/s
TF32 TFLOPS989
Form factorSXM5 / PCIe
Enquire
DGX System
NVIDIA DGX H200
The definitive AI supercomputer in a box. Eight H200 GPUs connected by NVLink with 640 GB aggregate memory.
GPUs8× H200 SXM
Total memory1.1 TB HBM3e
NVLink900 GB/s
Networking8× 400Gb InfiniBand
Enquire
Workstation GPU
NVIDIA RTX 6000 Ada
Professional visualisation and inference GPU. ECC memory, AV1 encode, and workstation-class reliability.
Memory48 GB GDDR6 ECC
CUDA Cores18,176
RT Cores3rd Gen
TDP300 W
Enquire
Data Processing Unit
NVIDIA BlueField-3
DPU for AI-powered cloud infrastructure. Offloads networking, storage and security from host CPUs.
Bandwidth400 Gb/s
Cores16× Arm A78
DRAM32 GB LPDDR5
PCIeGen 5
Enquire
Software Suite
NVIDIA AI Enterprise
End-to-end AI software platform. RAPIDS, Triton, NeMo, TAO Toolkit and enterprise support included.
FrameworksPyTorch, TF, JAX
InferenceTriton Server
MLOpsIncluded
Support24/7 Enterprise
Enquire
Server Platform
Dell PowerEdge XE9680
8-way GPU server optimised for AI and HPC. Supports up to eight SXM5 GPUs with NVLink Switch.
GPU slots8× SXM / PCIe
CPU2× Xeon Scalable
MemoryUp to 8 TB DDR5
NetworkingInfiniBand / Ethernet
Enquire
Why Tensor Works

Built for Serious Buyers

We exist to make enterprise AI infrastructure accessible to Australian organisations that can't afford to wait.

Dicker Data Partnership
Direct access to Australia's largest technology distributor, giving us priority allocation on high-demand GPU hardware.
100% Australian Owned
Australian-owned and operated. Compute stays onshore. Designed to meet government security and data sovereignty requirements.
Fast Fulfilment
Established supply chain relationships mean shorter lead times. We track global allocation and keep you informed throughout.
Dedicated Account Management
A named account manager for every engagement. No call centres — direct access to someone who knows your project.
Full-Stack Delivery
Hardware, software licences, networking fabric, and deployment support in a single engagement. One invoice, one relationship.
Government Panel Ready
Experienced with procurement frameworks used by federal and state government. Familiar with GITC, whole-of-government panels, and ICT procurement.
Contact

Talk to the Team

Tell us about your project and we'll come back with availability, lead times and indicative pricing — usually within one business day.


sam@vantyx.com.au
Melbourne, Victoria, Australia
Response within 1 business day
Thanks — your enquiry has been sent. We'll be in touch shortly.

Your information is kept confidential and never shared with third parties.