Delivery modes and operating boundary
AWS offers accelerated EC2 instances using GPUs and purpose-built accelerators; GPU instances can be combined with EKS, Batch and other AWS services. This page focuses on infrastructure-level compute rather than a single inference product.
EC2 gives control of instances and their operating system; orchestration and managed model services are separate AWS products with separate configuration, quotas and billing. GPU instance families serve different graphics, training, inference and HPC needs.
Computraq is prelaunch. Naming this provider here does not mean that Computraq hosts workloads on it, has tested its integration, or has a commercial relationship with it.
Buyer and procurement preparation
Use these questions to prepare a comparable, workload-specific evaluation. Product names describe the provider’s published offering; they do not establish availability for a particular account.
- Select accelerator family from framework support, memory, interconnect and workload type; confirm regional quota and on-demand/reserved/spot terms.
- Map supporting services such as storage, networking, Kubernetes or batch scheduling and include their costs and operational ownership.
- Review IAM, account boundaries, data location, network egress and service-specific contractual terms for the intended workload.
Developer and operator evidence
Before selecting capacity, create a small acceptance plan that records the model and revision, input shape, runtime, GPU configuration, concurrency, target latency or job completion, and recovery expectations. Provider documentation is evidence of product design, not proof that this workload has been tested.
- Use current instance specifications and compatible Deep Learning AMIs or container images; check driver/CUDA/framework versions.
- For multi-node workloads, validate placement, network bandwidth, collectives, checkpointing and retry behavior.
- Set budget/usage alerts and test interruption handling when using Spot capacity.