129 lines
4.7 KiB
Markdown
129 lines
4.7 KiB
Markdown
---
|
||
sidebar_position: 3
|
||
---
|
||
|
||
# Advanced Node Options
|
||
|
||
Beyond traditional hosting nodes, there are specialized **Edge Compute** and **Edge AI** configurations designed for enhanced earning potential and more demanding workloads.
|
||
|
||
## Edge Compute Node
|
||
|
||
**Based on GMKtec NUCBox M6 Ultra (Ryzen 5 7640HS)**
|
||
|
||
### Specifications
|
||
|
||
| Component | Specification |
|
||
|-----------|---------------|
|
||
| **CPU** | AMD Ryzen 5 7640HS (6 cores / 12 threads, up to ~5.0 GHz) |
|
||
| **RAM** | 32 GB |
|
||
| **Storage** | 1 TB NVMe SSD |
|
||
| **Power Usage** | ~30–40 W typical under load |
|
||
| **Network** | 2.5GbE LAN |
|
||
| **Form Factor** | Mini PC, silent and 24/7‑ready |
|
||
|
||
### Usage Options
|
||
|
||
#### Option 1: Use It Yourself (Private Edge)
|
||
|
||
- ✅ Full control of the hardware and software stack.
|
||
- ✅ Ideal for private AI, dev, or homelab workloads.
|
||
- ✅ No hosting fees – your own personal edge compute.
|
||
|
||
#### Option 2: Host for the Mycelium / ThreeFold Network
|
||
|
||
- ✅ Allocate "slices" of CPU, RAM, and storage as public capacity.
|
||
- ✅ Offer them on the marketplace and match customer demand.
|
||
- ✅ Earn SPORE with a guaranteed baseline via bid tiers.
|
||
- ✅ Example: standard compute tiers pay a minimum price (in SPORE) per slice.
|
||
|
||
#### Option 3: Hybrid
|
||
|
||
- ✅ Keep part of the machine for your private workloads.
|
||
- ✅ Rent out the remaining capacity to the network.
|
||
- ✅ Balance personal use and automated revenue.
|
||
|
||
---
|
||
|
||
## Edge AI Node
|
||
|
||
**Based on Ryzen AI MAX+ 395 platform**
|
||
|
||
### Specifications
|
||
|
||
| Component | Specification |
|
||
|-----------|---------------|
|
||
| **CPU** | AMD Ryzen™ AI MAX+ 395 (8C/16T, up to 5.1 GHz) |
|
||
| **AI Accelerator** | Integrated NPU (~50 TOPS) |
|
||
| **RAM** | 128 GB DDR5 |
|
||
| **Storage** | 2 TB NVMe SSD |
|
||
| **Power Usage** | ~45–55 W typical load |
|
||
| **Network** | Dual 2.5 G LAN |
|
||
| **Form Factor** | Compact, silent, designed for 24/7 AI workloads |
|
||
|
||
### AI‑Focused Features
|
||
|
||
- ✅ **Built‑in AI acceleration** – Dedicated NPU for local AI inference.
|
||
- ✅ **High memory headroom** – 128GB DDR5 for large models and context windows.
|
||
- ✅ **Edge AI ready** – Run models close to data sources for low latency.
|
||
- ✅ **Ideal for Mycelium Agents** and data‑sensitive AI workflows.
|
||
|
||
---
|
||
|
||
## How Rewards Flow
|
||
|
||
The earning logic for advanced nodes is the same as for any hosting node:
|
||
|
||
```mermaid
|
||
graph TD
|
||
A[Node Provides Capacity] --> B[User Rents Resources]
|
||
B --> C[80% to You<br/>SPORE Tokens]
|
||
B --> D[10% Burned<br/>Makes SPORE Scarcer]
|
||
B --> E[10% Mycelium<br/>Development]
|
||
|
||
C --> F[You Choose:<br/>Hold • Convert • Use]
|
||
D --> G[SPORE Supply Decreases<br/>Your Holdings More Scarce]
|
||
E --> H[Ecosystem Grows<br/>More Demand Over Time]
|
||
```
|
||
|
||
Advanced nodes simply position you for **higher‑value workloads** and **future AI‑centric demand**.
|
||
|
||
---
|
||
|
||
## Comparison: Edge vs Traditional Nodes
|
||
|
||
| Feature | Edge Compute | Edge AI | Traditional Mini Node |
|
||
|--------|--------------|--------|------------------------|
|
||
| **CPU Performance** | High (6‑core Zen 4) | Very High (8‑core AI) | Moderate |
|
||
| **RAM** | 32 GB | 128 GB DDR5 | 32 GB typical |
|
||
| **AI Capabilities** | Basic (CPU only) | Advanced NPU | None |
|
||
| **Power Usage** | ~30–40 W | ~45–55 W | ~60 W+ |
|
||
| **Best For** | General compute & cloud | AI/ML & agents | Entry‑level hosting |
|
||
|
||
---
|
||
|
||
## Getting Started with Edge Nodes
|
||
|
||
### Ready to Buy?
|
||
|
||
1. Research pricing and availability for **GMKtec NUCBox M6 Ultra** or **Ryzen AI MAX+ 395**.
|
||
2. Decide whether you want a **private**, **hosting**, or **hybrid** usage model.
|
||
3. Follow the **[Setup Guide](/hosters/setup)** to connect the node to your farm and the grid.
|
||
4. Read **[SPORE Economics](/hosters/economics)** to understand the reward model.
|
||
|
||
### Need Help Deciding?
|
||
|
||
- **Edge Compute** – Best for efficient general‑purpose compute with solid earning potential.
|
||
- **Edge AI** – Best for AI workloads, Mycelium Agents, and high‑end hosting.
|
||
- **Traditional nodes** – Best for starting small and learning the stack.
|
||
|
||
:::tip Pro Tip
|
||
Consider starting with a smaller, traditional hosting node to gain operational experience. Once you understand the economics and monitoring, upgrading to Edge Compute or Edge AI becomes a much easier decision.
|
||
:::
|
||
|
||
## Reference Links
|
||
|
||
- **GMKtec NUCBox M6 Ultra** – [gmktec.com/products/amd-ryzen-5-7640hs-mini-pc-nucbox-m6-ultra](https://www.gmktec.com/products/amd-ryzen-5-7640hs-mini-pc-nucbox-m6-ultra)
|
||
- **Ryzen AI MAX+ 395 EVO X2** – [gmktec.com/products/amd-ryzen%E2%84%A2-ai-max-395-evo-x2-ai-mini-pc](https://www.gmktec.com/products/amd-ryzen%E2%84%A2-ai-max-395-evo-x2-ai-mini-pc)
|
||
- **Learn SPORE Economics** – **[Farming/Hosting Economics](/hosters/economics)**
|
||
- **Community Discussion** – [t.me/threefoldfarmers](https://t.me/threefoldfarmers)
|