--- title: "CLOUD.NVMe compute instances" description: "High-performance cloud instances on U.2 NVMe (ZFS, ~200,000 IOPS) with AMD EPYC in a Polish data center. 99.99% SLA, resource scaling, API and automation. Test it free for 24 hours!" url: "https://nodea.io/en/cloud-nvme" lang: "en" updated: "2026-11-13" --- # CLOUD.NVMe compute instances ## CLOUD.NVMe compute instances — high performance The Performance line of instances on an array of U.2 NVMe drives on ZFS (~200,000 IOPS), powered by AMD EPYC Rome processors. Maximum IOPS and I/O throughput for demanding databases, in-memory caches and latency-sensitive workloads. Infrastructure across three availability zones in Poland — your data stays in the country. - [See pricing](#cennik) ## Why CLOUD.NVMe ### U.2 NVMe on ZFS, ~200,000 IOPS Up to ten times the IOPS of the SSD line. ZFS adds end-to-end checksums and snapshots, protecting data integrity under heavy write load. ### AMD EPYC Rome processors High single-core performance and plenty of cores — for databases, caches and high-request-rate applications that must stay responsive under peaks. ### Dedicated I/O, no sharing Local NVMe drives instead of SAN storage. No contention over disk resources with other tenants — predictable, low latency. ### Three availability zones Infrastructure spread across three independent data centers joined by a 100 Gbps fiber ring. A single facility failure does not stop your services. ### Monthly billing, no contracts You pay for the actual time an instance runs. No long-term commitments — scale vCPU, RAM and disk in step with the project. ### Data in Poland, GDPR-compliant Processing exclusively in Polish data centers, with Polish support and contracts. Full control over data location and reverse-DNS personalisation. ## What will you run on CLOUD.NVMe instances? ### Demanding databases ~200,000 IOPS on U.2 NVMe with ZFS handles intensive OLTP workloads, data warehouses and search indexes where every millisecond of write latency counts. ### In-memory caches and queues Redis, Memcached and message brokers get maximum I/O throughput and AMD EPYC cores — with no contention over disk resources with other tenants. ### High-traffic applications E-commerce stores, SaaS platforms and high-request-rate APIs stay responsive under peaks thanks to fast cores and low-latency storage. ## Integrated with the Nodea cloud platform ### WEB monitoring Every instance is connected to the app.nodea.io/app platform, where built-in monitoring of sites and services comes as standard. We watch availability and performance proactively, straight from the panel — with no separate tool to deploy. ### nodeaD — anti-DDoS with self-healing Our in-house anti-DDoS system, integrated with monitoring. It detects and blocks attacks proactively, and on failure it triggers self-healing — automatic response and recovery of services before users feel it. ### Connection Manager A cross-platform SSH and RDP agent integrated with app.nodea.io/app. Manage connections to every server from one place, without scattered configurations on individual workstations. *AI agents, API v1 and MCP* ## Run the app you built with an AI agent CLOUD.NVMe suits apps that need a fast disk: databases, APIs and busy online shops. An app you built with Claude Code, Codex, OpenCode or another popular harness (the tool that runs an AI agent) is easy to run on it. You order a CLOUD.NVMe instance in the panel with your SSH key and full root access, and your agent or CI pipeline deploys the app over SSH. From there the agent manages the server through REST API v1 and the MCP server: it restarts the instance, opens ports in the nodeaD agent firewall (if it runs on the instance), adds SSH keys, DNS records in zones hosted in Nodea DNS, and monitoring. It does so within the scopes of a token you create in the panel. [How to deploy from an AI agent](https://nodea.io/en/api-docs/ai-coding-agents) ## Three availability zones as standard ### Instances on a 3-AZ architecture You run instances across three independent availability zones (3-AZ) in the Tricity. This is an edge most VPS providers don't offer — you design for high availability at the infrastructure level. ### 100 Gbps fiber ring The three data centers are joined by a private 100 Gbps fiber ring. Replication and traffic failover between zones happen at low latency, within a single metropolitan area. ### Resilient to a zone failure A single facility failure does not stop your services — you keep replicas and backups in physically separated buildings. A natural foundation for a disaster recovery plan. ## CLOUD.NVMe instance pricing | Plan | vCPU | RAM | NVMe disk | Bandwidth | Transfer | Net (EUR) | Gross (EUR) | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | | CLOUD.NVME\_0.00 | 1 vCPU | 4 GB | 45 GB | 1 Gb/s | 2 TB | 7.99 EUR / month | 9.83 EUR / month | [Order now](https://app.nodea.io/app/compute/order?plan=cloud-nvme-0-00) | | CLOUD.NVME\_0.25 | 2 vCPU | 7.5 GB | 90 GB | 1 Gb/s | 2 TB | 15.99 EUR / month | 19.67 EUR / month | [Order now](https://app.nodea.io/app/compute/order?plan=cloud-nvme-0-25) | | CLOUD.NVME\_0.5 | 4 vCPU | 15 GB | 180 GB | 1 Gb/s | 2 TB | 30.99 EUR / month | 38.12 EUR / month | [Order now](https://app.nodea.io/app/compute/order?plan=cloud-nvme-0-5) | | CLOUD.NVME\_1 | 6 vCPU | 30 GB | 375 GB | 1 Gb/s | 2 TB | 61.99 EUR / month | 76.25 EUR / month | [Order now](https://app.nodea.io/app/compute/order?plan=cloud-nvme-1) | | CLOUD.NVME\_1.5 | 9 vCPU | 45 GB | 560 GB | 1 Gb/s | 2 TB | 92.99 EUR / month | 114.38 EUR / month | [Order now](https://app.nodea.io/app/compute/order?plan=cloud-nvme-1-5) | | CLOUD.NVME\_2 | 12 vCPU | 60 GB | 750 GB | 1 Gb/s | 2 TB | 122.99 EUR / month | 151.28 EUR / month | [Order now](https://app.nodea.io/app/compute/order?plan=cloud-nvme-2) | | CLOUD.NVME\_3 | 16 vCPU | 80 GB | 1,000 GB | 1 Gb/s | 2 TB | 163.99 EUR / month | 201.71 EUR / month | [Order now](https://app.nodea.io/app/compute/order?plan=cloud-nvme-3) | | CLOUD.NVME\_4 | 24 vCPU | 120 GB | 1,500 GB | 1 Gb/s | 2 TB | 245.99 EUR / month | 302.57 EUR / month | [Order now](https://app.nodea.io/app/compute/order?plan=cloud-nvme-4) | **Prices in PLN** | Plan | Net (PLN) | Gross (PLN) | | --- | --- | --- | | CLOUD.NVME\_0.00 | 33 PLN / month | 40.59 PLN / month | | CLOUD.NVME\_0.25 | 66 PLN / month | 81.18 PLN / month | | CLOUD.NVME\_0.5 | 132 PLN / month | 162.36 PLN / month | | CLOUD.NVME\_1 | 264 PLN / month | 324.72 PLN / month | | CLOUD.NVME\_1.5 | 396 PLN / month | 487.08 PLN / month | | CLOUD.NVME\_2 | 528 PLN / month | 649.44 PLN / month | | CLOUD.NVME\_3 | 704 PLN / month | 865.92 PLN / month | | CLOUD.NVME\_4 | 1,056 PLN / month | 1,298.88 PLN / month | - **CLOUD.NVME\_0.00**: Data centers: MP100 (Gdańsk), IQ (Gdańsk), GAC133 (Gdynia) - **CLOUD.NVME\_0.25**: Data centers: MP100 (Gdańsk), IQ (Gdańsk), GAC133 (Gdynia) - **CLOUD.NVME\_0.5**: Data centers: MP100 (Gdańsk), IQ (Gdańsk), GAC133 (Gdynia) - **CLOUD.NVME\_1**: Data centers: MP100 (Gdańsk), IQ (Gdańsk), GAC133 (Gdynia) - **CLOUD.NVME\_1.5**: Data centers: MP100 (Gdańsk), IQ (Gdańsk), GAC133 (Gdynia) - **CLOUD.NVME\_2**: Data centers: MP100 (Gdańsk), IQ (Gdańsk), GAC133 (Gdynia) - **CLOUD.NVME\_3**: Data centers: MP100 (Gdańsk), IQ (Gdańsk), GAC133 (Gdynia) - **CLOUD.NVME\_4**: Data centers: MP100 (Gdańsk), IQ (Gdańsk), GAC133 (Gdynia) Net prices; 23% VAT is added. ### Availability backed by a 99.99% SLA We guarantee 99.99% service availability over the year, written into an SLA. Every instance is monitored with Zabbix, and alerts are handled by a technical team made up exclusively of Linux administrators — often before you notice a problem. ### Try an instance free for 24 hours We're confident in the quality of our services, so we let you test them first. Just note in the order that you'd like a trial and we'll spin up your instance free of charge for 24 hours — no strings attached. ## We'll migrate you in for free Every client with a Linux administration plan is migrated free of charge, with no billing for extra admin hours. In our experience a well-executed migration is an investment that pays off in years of successful cooperation. [See administration plans](https://nodea.io/en/administracja-serwerami-linux) ## Frequently asked questions ### How does NVMe differ from the SSD line? CLOUD.NVMe is built for maximum performance: U.2 NVMe drives on ZFS reach around 200,000 IOPS, many times more than the SSD arrays (~20,000 IOPS) of the CLOUD.SSD line. It's the choice for latency-sensitive workloads — intensive databases, in-memory caches and high-request-rate applications. When load is more moderate and price-to-performance matters, CLOUD.SSD is enough. ### Why ZFS under NVMe? ZFS adds end-to-end checksums, snapshots and mechanisms that guard against silent data corruption. Under heavy write load you get both the high performance of NVMe and a higher level of data integrity. ### How does billing work? Instances are billed monthly, with no long-term contracts and no cancellation penalties. You pay for the period an instance runs and scale resources in step with your project. ### Does my data stay in Poland? Yes. All instances run in Polish data centers across three availability zones in the Tricity, with Polish support and contracts — and in compliance with GDPR. ### Can you administer the server for me? Yes. We offer full Linux administration in four plans — with Zabbix monitoring, proactive incident response, security audits, DR tests and backup storage. We act on monitoring alerts ourselves, so you don't have to report outages. ### Can I deploy from Claude Code, Codex or CI? Yes. The instance comes with full root access and the SSH key you chose when ordering, so a coding agent (Claude Code, Codex, OpenCode and other popular harnesses) or a CI pipeline deploys the app over SSH. Through REST API v1 and the MCP server (https://app.nodea.io/mcp, a personal token from the panel) the agent restarts and stops the instance, opens ports in the nodeaD agent firewall if it runs on the instance, and manages SSH keys and records in Nodea DNS zones. You order the instance in the panel. The step-by-step setup is in the [guide to deploying from AI coding agents](https://nodea.io/en/api-docs/ai-coding-agents). ## Let's talk about your project [Contact us](https://nodea.io/en/kontakt) --- Languages: [pl](https://nodea.pl/pl/cloud-nvme.md) Site index: [llms.txt](https://nodea.io/en/llms.txt)