GPU servers
NVIDIA HGX H100 8-GPU server
8 x H100 SXM, the proven training workhorse.
HGX H100 is the most widely deployed AI training platform of its generation. Eight H100 SXM GPUs with 80 GB of HBM3 each give 640 GB of GPU memory per node, linked by NVLink. New and certified used nodes are both available on request.
Built for
- Model training
- Cost-efficient inference
- Expanding existing H100 clusters
Questions
No. Prices for AI hardware move weekly, so every order gets a landed AUD quote: product, freight, insurance, customs clearance and GST shown line by line. Send your list and we quote against confirmed supply.