1. Service Description
Cloud Compute provides on-demand GPU and CPU virtual machines for AI/ML workloads, rendering, scientific computing, and general-purpose compute.
2. Infrastructure
2.1. GPU Instances:
- NVIDIA H100, A100, L40S, and RTX series GPUs
- Optimized for AI training, inference, and CUDA workloads
- NVLink and NVSwitch for multi-GPU configurations
2.2. CPU Instances:
- AMD EPYC and Intel Xeon processors
- High-memory and compute-optimized configurations
- Bare metal options for maximum performance
2.3. Storage:
- NVMe SSD local storage
- Network-attached block storage
- Object storage (S3-compatible)
3. Locations
Instances are available in:
- Australia: Sydney, Melbourne
- International: Available through our partner network (US, EU, APAC)
4. Billing
4.1. Hourly Billing: Pay for what you use, billed per hour (minimum 1 hour).
4.2. Reserved Instances: Commit to 1, 6, or 12 months for significant discounts.
4.3. Spot Instances: Discounted rates for interruptible workloads (where available).
4.4. Storage: Charged separately based on GB provisioned and data transfer.
5. Service Level Agreement (SLA)
- Target Availability: 99.9% for individual instances.
- Redundancy: Multi-region deployments recommended for high availability.
Rebates: If we fail to meet our availability target, credits will be applied to your account based on the duration of the outage.
Note: Scheduled maintenance, customer-initiated actions, and force majeure are excluded.
6. Data & Security
6.1. Isolation: Each instance runs in an isolated environment. Multi-tenancy is handled at the hypervisor level.
6.2. Encryption: Data at rest can be encrypted. Data in transit is encrypted via TLS.
6.3. Backups: You are responsible for backing up your data. We provide snapshot functionality.
6.4. Termination: When you terminate an instance, data on local storage is permanently deleted.
7. Acceptable Use
7.1. Prohibited Uses:
- Cryptocurrency mining (unless on dedicated reserved instances with prior approval)
- DDoS attacks or network abuse
- Distribution of malware or illegal content
- Violation of export controls
7.2. Resource Abuse: We reserve the right to suspend instances that negatively impact other customers or our infrastructure.
8. Support
8.1. Self-Service: Provision, manage, and terminate instances via the portal or API.
8.2. Technical Support: Available for infrastructure issues. Application-level support is your responsibility.
8.3. Professional Services: Custom configurations, architecture reviews, and managed services available on request.