Compute C2 GPU virtual servers are designed for the demanding needs of data analytics, machine learning, and other GPU-accelerated workloads. High-frequency CPUs coupled with NVIDIA A100 GPUs provide the ideal platform to crunch vast amounts of data into insights and models.
Compute C2 GPU
High-performance virtual servers with data processing acceleration
Compute C2 GPU Features
High-Frequency Intel Xeon Processors
Powered by high frequency Intel Xeon processors, up to 3.8GHz (compared with 2.2GHz for our normal virtual servers), maximising data processing capacity
All resources of the C2 GPU virtual servers are dedicated and reserved, ensuring that performance is always available and predictable
NVIDIA A100 GPUs
Each virtual server features slices of NVIDIA A100 80GB GPUs, each slice having dedicated GPU compute pipelines and 20GB of VRAM.
Billed by the hour
No upfront commitment or expensive servers to buy, consume a GPU for an hour and only pay for that hour.
All Compute C2 GPU virtual servers have access to 25Gbit/s networking and advanced networking capabilities such as Security Groups.
Prices for Compute C2 GPU virtual servers using community supported Linux images, images you have uploaded into Catalyst Cloud, and any images with "Bring Your Own Support" license terms.
|GPU Compute Pipelines
|Price per hour (ex GST)
|Price per month (ex GST)
- Monthly prices are based on an approximation of 730 hours per month. Actual price charged each month depends on the number of hours in the month.
- All licenses required are your responsibility to obtain and comply with where needed, whether the image is provided by Catalyst Cloud or you have uploaded an image.
- Windows is not currently supported on these virtual servers. NVIDIA officially support Ubuntu only for driver support in these virtual servers.