> ## Documentation Index
> Fetch the complete documentation index at: https://docs.nvisy.com/llms.txt
> Use this file to discover all available pages before exploring further.

# System Requirements

> Hardware and software requirements for on-premises deployment

<Warning>
  The hardware figures on this page are **illustrative starting points**, not
  measured requirements. Concurrency, IOPS, throughput, latency, and GPU
  speed-up numbers have not been benchmarked against Nvisy workloads — they
  describe generic system capacity. Size your deployment from a trial run with
  your own documents, formats, and detection settings.
</Warning>

## Production Requirements

| Component             | Requirement                                                 |
| --------------------- | ----------------------------------------------------------- |
| **Operating System**  | Ubuntu 22.04 LTS or RHEL 9                                  |
| **CPU**               | 16+ cores with AVX2 support                                 |
| **RAM**               | 32 GB+                                                      |
| **Storage**           | 500 GB+ NVMe SSD                                            |
| **GPU**               | NVIDIA GPU with 8GB+ VRAM (optional, for faster processing) |
| **Network**           | 10 Gbps with load balancer                                  |
| **Container Runtime** | Docker 20.10+ or Kubernetes 1.21+                           |
| **High Availability** | 3+ nodes for redundancy                                     |

## Operating System Support

### Supported Distributions

<Tabs>
  <Tab title="Ubuntu">
    * Ubuntu 20.04 LTS (Focal)
    * Ubuntu 22.04 LTS (Jammy)
    * Ubuntu 24.04 LTS (Noble)

    **Recommended:** Ubuntu 22.04 LTS
  </Tab>

  <Tab title="Red Hat">
    * RHEL 8.x
    * RHEL 9.x
    * Rocky Linux 8.x, 9.x
    * AlmaLinux 8.x, 9.x

    **Recommended:** RHEL 9 or Rocky Linux 9
  </Tab>

  <Tab title="Debian">
    * Debian 11 (Bullseye)
    * Debian 12 (Bookworm)

    **Recommended:** Debian 12
  </Tab>

  <Tab title="Other">
    * Amazon Linux 2023
    * SUSE Linux Enterprise Server 15

    Other distributions may work but are not officially supported.
  </Tab>
</Tabs>

## Hardware Requirements

### CPU

* 16+ cores (32+ threads)
* Intel Xeon or AMD EPYC
* AVX2 instruction set support
* 3.0 GHz+ base clock

For high-volume processing:

* 32+ cores
* Latest generation processors
* AVX-512 support

### Memory (RAM)

* **32 GB** - A reasonable starting point for moderate concurrency
* **64 GB+** - For high-volume deployments

Actual memory per job depends on document size, format, and whether OCR runs,
so treat concurrency figures as illustrative until you have measured your own
workload.

### Storage

* **500 GB NVMe SSD** - Recommended starting point
* **1+ TB NVMe SSD** - For high-volume or long-term log retention
* RAID 10 configuration recommended for production

### Storage Performance

| Requirement    | Minimum  | Recommended |
| -------------- | -------- | ----------- |
| **IOPS**       | 3,000    | 10,000+     |
| **Throughput** | 125 MB/s | 500 MB/s+   |
| **Latency**    | \< 10ms  | \< 1ms      |

### GPU (Optional)

GPU acceleration significantly improves processing speed:

**Supported GPUs:**

* NVIDIA T4 (16 GB)
* NVIDIA A10 (24 GB)
* NVIDIA A100 (40/80 GB)
* NVIDIA L4 (24 GB)
* NVIDIA L40 (48 GB)

**Requirements:**

* CUDA 11.8 or higher
* NVIDIA driver 520.61.05 or higher
* Compute capability 7.0+

**Performance Improvement:**

* A GPU can substantially reduce analysis time for model-driven detection and
  OCR-heavy documents. The size of the improvement depends on the model,
  document mix, and CPU baseline, so benchmark before sizing around it.
* Recommended for high-volume deployments

## Network Requirements

### Bandwidth

* **10 Gbps** - Recommended for production
* **10+ Gbps with redundancy** - For high-volume enterprise deployments

Bandwidth needs scale with document size and how much traffic moves between the
service and remote object storage.

### Ports

Required network ports:

| Port | Protocol | Purpose    | Required          |
| ---- | -------- | ---------- | ----------------- |
| 8080 | TCP      | HTTP API   | Yes               |
| 8443 | TCP      | HTTPS API  | Recommended       |
| 9090 | TCP      | Metrics    | Optional          |
| 5432 | TCP      | PostgreSQL | Yes (if external) |
| 4222 | TCP      | NATS       | Yes (if external) |

### Firewall Rules

Allow inbound traffic:

```bash theme={null}
# HTTP API
sudo ufw allow 8080/tcp

# HTTPS API
sudo ufw allow 8443/tcp

# Metrics (from monitoring network only)
sudo ufw allow from 10.0.0.0/8 to any port 9090
```

## Software Dependencies

### Container Runtime

<Tabs>
  <Tab title="Docker">
    **Required Version:** 20.10+

    **Installation:**

    ```bash theme={null}
    # Ubuntu/Debian
    curl -fsSL https://get.docker.com -o get-docker.sh
    sudo sh get-docker.sh

    # Verify
    docker --version
    ```

    **Docker Compose:**

    ```bash theme={null}
    # Required version: 2.0+
    sudo apt-get install docker-compose-plugin
    docker compose version
    ```
  </Tab>

  <Tab title="Kubernetes">
    **Required Version:** 1.21+

    **Recommended:** 1.27+

    **Supported Distributions:**

    * Vanilla Kubernetes
    * Amazon EKS
    * Google GKE
    * Azure AKS
    * Red Hat OpenShift 4.10+
    * Rancher

    **Required Components:**

    * kubectl CLI
    * Helm 3.0+
    * Persistent volume provisioner
  </Tab>

  <Tab title="Podman">
    **Supported:** Yes (experimental)

    **Version:** 4.0+

    **Note:** Limited testing, use Docker for production
  </Tab>
</Tabs>

### Database

**PostgreSQL** is required for metadata and audit logs:

**Version:** 12+
**Recommended:** 14+

**Storage Requirements:**

* 10 GB minimum
* 50 GB recommended
* Scales with audit log retention

### Messaging

**NATS** is required for job messaging between services:

**Version:** 2.9+

Memory needs scale with queued job volume; a small allocation is sufficient for
typical deployments.

## High Availability Requirements

For production HA deployment:

### Compute

* **Minimum:** 3 nodes
* **Recommended:** 5+ nodes
* **Load balancer:** NGINX, HAProxy, or cloud LB

### Database

* **PostgreSQL:** Primary + 2 replicas
* **Replication:** Streaming replication
* **Automatic failover:** Patroni or similar

### Messaging

* **NATS:** Clustered (3+ nodes) for redundancy

### Storage

* **Shared storage:** NFS, S3, or similar
* **Replication:** 3x minimum
* **Backup:** Daily with off-site replication

## Cloud Provider Recommendations

### AWS

**Instance Types:**

* **Minimum:** t3.xlarge (4 vCPU, 16 GB)
* **Recommended:** c6i.4xlarge (16 vCPU, 32 GB)
* **With GPU:** g5.2xlarge (1x A10G GPU)

**Storage:** EBS gp3 with 10,000 IOPS

### Azure

**VM Sizes:**

* **Minimum:** Standard\_D4s\_v5 (4 vCPU, 16 GB)
* **Recommended:** Standard\_D16s\_v5 (16 vCPU, 64 GB)
* **With GPU:** Standard\_NC8as\_T4\_v3 (1x T4 GPU)

**Storage:** Premium SSD

### Google Cloud

**Machine Types:**

* **Minimum:** n2-standard-8 (8 vCPU, 32 GB)
* **Recommended:** n2-standard-16 (16 vCPU, 64 GB)
* **With GPU:** n1-standard-8 + 1x T4

**Storage:** SSD persistent disk

## Capacity Planning

### Concurrent Jobs

Estimate required resources based on workload:

These pairings are illustrative and unbenchmarked. The "concurrent jobs" column
is a planning target, not a measured result — the worker-count guidance in
[Performance Tuning](/deployment/on-premise/getting-started) is a conservative
CPU-bound floor and will suggest a lower number for the same hardware.

| Concurrent Jobs | CPU Cores | RAM    | Storage IOPS |
| --------------- | --------- | ------ | ------------ |
| 10              | 8         | 16 GB  | 3,000        |
| 25              | 16        | 32 GB  | 10,000       |
| 50              | 32        | 64 GB  | 20,000       |
| 100             | 64        | 128 GB | 40,000       |

### Storage Sizing

Calculate storage needs:

```
Total Storage = (Avg Document Size × Daily Documents × Retention Days) + Base Storage

Example:
(5 MB × 1000 documents × 30 days) + 50 GB = 200 GB
```

## Benchmarking

Test your environment:

```bash theme={null}
# CPU benchmark
sysbench cpu --threads=16 --time=60 run

# Memory benchmark
sysbench memory --memory-total-size=100G run

# Disk I/O benchmark
fio --name=randwrite --ioengine=libaio --rw=randwrite \
    --bs=4k --numjobs=4 --size=4g --runtime=60

# Network benchmark
iperf3 -c <server-ip> -t 60 -P 4
```

## Next Steps

<CardGroup cols={2}>
  <Card title="Installation" icon="download" href="/deployment/on-premise/installation">
    Install Nvisy on-premises
  </Card>

  <Card title="Getting Started" icon="rocket" href="/deployment/on-premise/getting-started">
    Configure and start using Nvisy
  </Card>
</CardGroup>