--- title: "TVE (POC) Infrastructure Requirements" slug: "tve-poc-infrastructure-requirements" updated: 2026-07-28T13:21:17Z published: 2026-07-28T13:21:17Z canonical: "help.hyperscience.ai/tve-poc-infrastructure-requirements" --- > ## Documentation Index > Fetch the complete documentation index at: https://help.hyperscience.ai/llms.txt > Use this file to discover all available pages before exploring further. # TVE (POC) Infrastructure Requirements Hyperscience can be installed on Linux VMs in an on-premises instance or private cloud using Docker or Podman containers. The platform is accessible through a web application and API. This article lists the infrastructure requirements for a Technical Validation Event (TVE), a short technical validation or proof of concept that uses small volumes of files, such as 1,000 files. A TVE typically runs on a **single application VM** with the Application DB and File Store located locally on that VM. The standard installation package includes a PostgreSQL container, so you do not need to provision an external database. > [!NOTE] > This local, single-VM deployment is intended for evaluation only > > It is not recommended for production because it does not support High Availability / Disaster Recovery (HA/DR). For production requirements, see [Infrastructure Requirements.](/deployment/docs/infrastructure-requirements) ## Specifications > [!WARNING] > A TVE is always deployed on the latest Hyperscience release > > This article lists requirements for the latest version only. For requirements for an earlier versions, see the corresponding section of the [Infrastructure Requirements](/deployment/docs/infrastructure-requirements) article. ### Internet browsers Hyperscience supports the latest versions of Google Chrome and Microsoft Edge. For the best possible user experience, we recommend browser dimensions of at least 1280 x 720 pixels. ## Application server A TVE runs on a single application VM. The requirements below apply to that VM. If you plan to process Semi-structured documents, you may also provision an optional Trainer VM (see [Trainer (optional)](/deployment/docs/tve-poc-infrastructure-requirements#trainer)). ### Operating system The table below lists the supported operating systems. | Operating system | Supported versions | | --- | --- | | Ubuntu | 22.04 and 24.04 | | RHEL | 8.10, 9.8, 9.10, and 10 | ### Container instances The table below lists the supported container instances for each operating system. | Operating system | Supported container instance | | --- | --- | | RHEL | Podman 3.3.1 and later | | Ubuntu 22.04 and 24.04 | Docker 25.0.4 and later | Note the following: - The container instance must be installed on the VM and set to autostart, with no containers running before installation. - The container instance’s preferred storage driver is *overlay2*. - The container instance can be called `docker-latest` or `docker` if you are using Docker, or `podman` if you are using Podman. ### Local storage Because a TVE stores both the Application DB and the File Store locally, the VM's local storage must accommodate the application, the database, and processed files. The requirements below are in addition to the space consumed by the OS. | Total local storage | Volume for downloading, extracting, and deploying the application (typically the root ( / ) volume) | Volume Docker or Podman is set up to use for the application image (typically located at /var) | | --- | --- | --- | | At least 400 GB | At least 300 GB | At least 100 GB | > [!NOTE] > To maximize the amount of available space, we recommend keeping no more than two bundles in your local storage. Additional space is recommended to accommodate the local File Store and future expansion. ### CPU - Intel x86_64 is a requirement. ARM is not supported. - The system requires a minimum of 8 CPU cores in the VM. > [!NOTE] > **Definition of "CPU cores":** In this article, "CPU cores" refers to threads on Intel processors with enabled Hyper-Threading, and virtual CPUs (vCPUs) on cloud compute instances (for example, AWS, GCP). For example, a VM using 4 physical cores on an Intel processor with Hyper-Threading enabled has 8 logical cores (8 threads) and is considered to have 8 cores. An AWS EC2 or GCE instance with 8 vCPUs is considered to have 8 cores. ### RAM The system requires a minimum of 32 GB of RAM in the VM. For optimal performance, we recommend a 1:4 ratio of CPU cores to gigabytes of RAM (for example, an 8-core VM should have 32 GB of RAM). > [!NOTE] > **Burstable-performance machines are not supported. Such machines are:** > > - AWS: T-series. To learn more, see Amazon’s [Burstable performance instances](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/burstable-performance-instances.html). > - Azure: B-series. To learn more, see Microsoft’s [B-series burstable virtual machine sizes](https://docs.microsoft.com/en-us/azure/virtual-machines/sizes-b-series-burstable). > - Google Cloud: shared core. To learn more, see Google’s [General-purpose machine family](https://cloud.google.com/compute/docs/general-purpose-machines). > [!NOTE] > **Servers may consume 100% of machine resources for prolonged periods of time.** > > - For virtualized images, the expectation is that a vCPU would be equivalent to a physical CPU core in understanding capacity requirements for our platform. > - The nature of burstable-performance machines does not allow them to constantly utilize 100% of the CPUs’ resources, which results in system slowness. ## Application database In a TVE, the recommended option is the **PostgreSQL container included in the installation package**. If you use it, you do not need to provision an external database. If you prefer to connect the TVE to an external database, Hyperscience supports the same options as a production deployment. For the full list of supported database versions (PostgreSQL, Oracle, and Microsoft SQL Server) and their privilege requirements, see the [Infrastructure Requirements](https://help.hyperscience.ai/deployment/docs/infrastructure-requirements) article. > [!NOTE] > PostgreSQL and MSSQL are the recommended database options. Migrating existing data between different database types is not supported. ## File storage In a TVE, the File Store is located locally on the application VM, so no external file store is required. Ensure the local storage accounts for the volume of files you plan to process (see [Local storage](/deployment/docs/tve-poc-infrastructure-requirements#local-storage)). If you prefer to use an external file store, the following options are supported: - AWS S3 bucket - Azure Blob Storage - Google Cloud Storage - A networked file store (such as NFS or CIFS) ## Trainer (optional) The Hyperscience Trainer runs separately from the main application and communicates with it via the API. The Trainer supports select long-running tasks and very large file downloads and uploads that might otherwise negatively impact document processing time. A Trainer is **required to process Semi-structured documents** (for example, invoices, checks, paystubs, and bills). If your TVE only processes Structured documents (forms), a Trainer is not required. When used, the Trainer should run on its own VM. ### Operating system and container instance Trainer VMs run on the same infrastructure as the application server. The operating system and container instance requirements described above also apply to the Trainer VM. ### Database The Trainer uses its own local PostgreSQL database, which comes pre-packaged with the Hyperscience bundle. You do not need to provision an external database for the Trainer VM. ### Storage The Trainer VM has the same local storage requirements as the application server. Use local storage with the Trainer; do not use shared storage, as this may cause data to be overwritten and training jobs to fail. ### VM CPU cores We require 16 CPU cores in the Trainer VM for training Semi-structured models. ### RAM The Trainer needs 4 GB of RAM for each CPU core in the Trainer VM, with a minimum total of 64 GB of RAM. For a 16-core CPU, this means 64 GB of RAM for maximum performance. ## Installation overview Our software bundle is provided as a single, easy-to-install tarball. Sample installation instructions can be found in [TVE (POC) Installation Process](https://help.hyperscience.ai/deployment/docs/tve-poc-installation-process). We also help our customers install over screenshare. The bundle is typically delivered over SFTP, or via another file transfer option of your choice. A trial phase where a potential customer tests the platform with clear success criteria. It’s designed to show that the product works well for their needs. This phase ensures that both sides are aligned before moving forward. A separate machine dedicated to handling resource-heavy tasks like training Identification models. It operates independently and connects to the main application through the API.