Node Tiers Explained
Understanding Cumulus, Nimbus, and Stratus tiers — collateral, specs, and rewards.
Node Tiers Explained
The Flux network uses a three-tier node system to accommodate operators with different levels of hardware and investment. Each tier has specific minimum hardware requirements and a corresponding FLUX collateral amount. Higher-tier nodes provide more computing resources to the network and, in return, earn a larger share of block rewards. The three tiers are Cumulus, Nimbus, and Stratus.
This tiered architecture allows the Flux network to offer a diverse range of compute resources to application developers. Lightweight apps can be deployed on Cumulus-class nodes, while resource-intensive workloads like AI inference, large databases, or high-traffic web applications can target Nimbus or Stratus nodes for more powerful hardware.
Tier Comparison
| Specification | Cumulus | Nimbus | Stratus |
|---|---|---|---|
| Collateral Required | 1,000 FLUX | 12,500 FLUX | 40,000 FLUX |
| CPU Cores (min) | 2 cores | 4 cores | 8 cores |
| RAM (min) | 8 GB | 32 GB | 64 GB |
| SSD Storage (min) | 220 GB | 440 GB | 880 GB |
| Bandwidth (min) | 30 Mbps | 60 Mbps | 100 Mbps |
| CPU Threads (min) | 4 threads | 8 threads | 16 threads |
| Reward Share (of node pool) | ~7.5% | ~12.5% | ~80% |
The specifications listed above are the minimum requirements to register a node on each tier. In practice, Flux strongly recommends exceeding these minimums — especially for RAM and SSD — because FluxOS reserves a portion of resources for the system itself, and hosted applications compete for the remaining capacity. Nodes that are consistently resource-constrained may fail benchmark tests and lose their position in the reward queue.
Cumulus — Entry-Level Tier
Cumulus is the entry-level tier, designed for operators who want to participate in the Flux network with a modest hardware and capital commitment. With a collateral requirement of just 1,000 FLUX, it is the most accessible way to become a node operator. Cumulus nodes provide 2 CPU cores, 8 GB of RAM, and 220 GB of SSD storage to the network.
Cumulus nodes are well-suited for hosting lightweight applications such as static websites, small APIs, lightweight blockchain full nodes, and microservices. Because many operators run Cumulus nodes, the per-node reward is smaller, but the lower investment makes it an excellent starting point for newcomers who want to learn how the system works before scaling up.
A recommended Cumulus setup is a dedicated mini PC or a small server with at least 4 CPU cores, 16 GB of RAM, and a 500 GB NVMe SSD. This gives comfortable headroom above the minimums and ensures your node passes periodic benchmarks without issues.
Nimbus — Mid-Level Tier
Nimbus occupies the middle ground, requiring 12,500 FLUX in collateral and providing significantly more compute resources: 4 CPU cores, 32 GB of RAM, and 440 GB of SSD storage with at least 60 Mbps of bandwidth. Nimbus nodes handle medium-weight workloads including content management systems, medium-traffic web applications, databases, and blockchain validators.
The economics of running a Nimbus node are attractive because the reward share per node is higher than Cumulus, and there are fewer Nimbus nodes competing for the Nimbus reward pool. Operators who have the hardware and capital often find that Nimbus offers the best balance between investment cost and reward yield.
Stratus — High-Performance Tier
Stratus is the premium tier, designed for operators with enterprise-grade hardware. It requires 40,000 FLUX in collateral and demands 8 CPU cores, 64 GB of RAM, 880 GB of SSD storage, and at least 100 Mbps of bandwidth. Stratus nodes handle the most demanding workloads on the network, including AI/ML inference, large-scale databases, high-traffic dApps, and GPU-accelerated computing tasks.
Stratus nodes receive the largest share of the node reward pool — approximately 80% of all node rewards are allocated to the Stratus tier. Despite the significant upfront investment, the per-node reward for Stratus is substantially higher than the other tiers, making it the preferred choice for serious operators looking to maximize their return.
Resource Allocation & FluxOS Overhead
It is important to understand that not all of a node's hardware resources are available for hosting applications. FluxOS itself reserves a portion of system resources for operating system processes, the Flux daemon, MongoDB (used for local state management), Docker overhead, and benchmark verification tasks. The approximate reserved resources are:
| Resource | FluxOS Reserved (approx.) |
|---|---|
| CPU | ~0.5 cores |
| RAM | ~2 GB (Cumulus) / ~3 GB (Nimbus) / ~4 GB (Stratus) |
| SSD | ~50 GB for OS, daemon, and databases |
| Bandwidth | Variable — benchmark tests and daemon sync use bandwidth periodically |
This means a Cumulus node with exactly 8 GB of RAM will have roughly 6 GB available for hosting apps, while a Stratus node with 64 GB will have approximately 60 GB available. This is another reason why exceeding the minimum hardware specifications is strongly recommended.
Choosing the Right Tier
- 1
Budget & Capital
If you are just starting out and want minimal financial exposure, Cumulus is the safest choice. As you grow more confident and accumulate more FLUX, you can upgrade to higher tiers.
- 2
Available Hardware
Evaluate the hardware you already own. If you have a powerful server sitting idle, it may qualify for Nimbus or even Stratus. Running a tier below your hardware's capability means leaving potential rewards on the table.
- 3
Reward Goals
Higher tiers earn proportionally more rewards. If maximizing FLUX income is your primary goal and you have the capital, Stratus offers the highest per-node payout. However, running multiple Cumulus nodes can sometimes match or exceed a single Stratus node in total rewards, depending on the current node counts per tier.
- 4
Network Demand
Consider what types of applications are being deployed. As AI workloads grow, demand for Stratus-class resources increases, potentially leading to greater utilization bonuses and additional revenue streams in the future.
You can check the current number of active nodes per tier on the Flux dashboard at home.runonflux.io. This information is useful for estimating your expected reward frequency and APY before committing collateral.
Sources & Further Reading
Other articles in Getting Started
What is Flux?
An introduction to the Flux blockchain, decentralized compute network, and its ecosystem.
Collateral Requirements
How collateral works, locking mechanisms, and investment considerations.
Rewards & APY Explained
How block rewards are distributed, APY calculations, and reward projections.
Provider Registration Guide
Step-by-step guide to registering as a hosting provider on the platform.