Provision Compute Resources
Provision Compute Resources
Cloud providers don't make you configure a VM from nothing. They offer preconfigured instance types tuned to a role — general purpose, compute optimized, memory optimized, and a handful of specialists — and the defaults work for most situations.
Provision Compute Resources
CSPs provide preconfigured VM instances to simplify deployment. Each instance bundles CPU, GPU, memory, storage, and network settings tuned to a specific role. They're customizable, but the defaults will be fine for most situations. AWS offers several instance types — tap one to see what it's built for.
Sample Instance Selection: Storage Optimized
AWS D3 instance types are optimized for storage workloads — high I/O throughput built on HDD arrays rather than sheer processor power. Click a row to see it sized against the others.
| Name | vCPU | Memory (GiB) | Storage | Storage type | Network bandwidth |
|---|---|---|---|---|---|
| d3.xlarge | 4 | 32 | 3×2 TB | HDD | Up to 15 Gbps |
| d3.2xlarge | 8 | 64 | 6×2 TB | HDD | Up to 15 Gbps |
| d3.4xlarge | 16 | 128 | 12×2 TB | HDD | Up to 15 Gbps |
Resizing After Deployment
The instance you pick on day one isn't permanent. Compute resources can be adjusted manually after deployment — but that flexibility comes with its own operational cost.
Why you'd resize
Misconfigurations, unanticipated growth, or a push to optimize costs are the usual triggers. Cloud administrators regularly work with resource allocation, resource limits, and sizing issues throughout a VM's life — provisioning is rarely a one-time decision.
What it costs you
Typically, the instance must be offline during the resize process. That downtime has to be planned for, and getting the match between workload needs and instance type wrong in either direction can hit cost, performance, or availability.
Multi-Threading
Multi-threading is a system's ability to divide tasks among multiple CPU cores, letting several functions run simultaneously. Provisioning vCPUs means calculating cores and threads together. Adjust the sliders to see the AWS formula in action.