Becoming a Provider with FluxCore
Install FluxCore, benchmark your GPU, join the marketplace, and earn from rentals with auto-switch mining fallback.
Becoming a FluxEdge Provider with FluxCore
FluxCore is the desktop application that connects your GPU hardware to the FluxEdge marketplace. By installing FluxCore, you can monetize idle GPU time by renting your machine to users who need compute for AI, rendering, and other intensive tasks. FluxCore also supports native FLUX mining and NiceHash integration as fallback income when your machine is not rented.
System Requirements
- β’GPU: NVIDIA GPU required (AMD support for GPU rentals is limited). Check the supported NVIDIA drivers list.
- β’OS: Windows, Linux, or macOS
- β’Drivers: NVIDIA 570 series drivers recommended. CUDA toolkit 12.8 required on Linux.
- β’Network: Stable internet connection with good upload/download speeds
- β’Windows specific: WSL (Windows Subsystem for Linux) and virtualization enabled in BIOS
Linux Installation
- 1
Verify NVIDIA driver compatibility
Check the FluxCore Supported NVIDIA Drivers page. Install the recommended 570 series driver.
- 2
Install CUDA toolkits
Install both cuda-toolkit-12-8 AND nvidia-cuda-toolkit packages BEFORE installing FluxCore.
- 3
Run the installer
Execute the FluxCore setup script and follow the prompts.
- 4
Install prerequisites
Select option 1 to install prerequisites when prompted. Enter your sudo password.
FluxCore Linux installation
# Install CUDA prerequisites first
sudo apt install -y cuda-toolkit-12-8 nvidia-cuda-toolkit
# Run the FluxCore installer
bash -i <(curl -s https://download.fluxcore.ai/setup.sh)Windows Installation
- 1
Enable virtualization
Enter BIOS and enable Intel VT-x or AMD-V virtualization. This is required for WSL.
- 2
Download FluxCore
Download the Windows installer from the FluxCore website.
- 3
Run the installer
FluxCore will auto-install WSL, CUDA, and TensorFlow dependencies as needed.
- 4
Save your credentials
During "Register Machine," download the Machine ID and Password file. This enables remote management.
Joining the FluxEdge Marketplace
After installation, you need to join a Kubernetes cluster to make your machine available for rent on FluxEdge.
- 1
Install the Service Module
Navigate to Account > Modules in FluxCore, or use the menu bar shortcut.
- 2
Subscribe to service categories
Choose which workload types to accept: AI, Rendering, Cloud Gaming, etc.
- 3
Select a cluster server
Choose an available Kubernetes cluster in your geographic region for optimal latency.
- 4
Accept terms and join
FluxCore installs Kubernetes software, joins the cluster, and your machine appears on the marketplace.
Benchmarking
FluxCore includes integrated benchmarking tools that evaluate your hardware across three dimensions: AI Training, AI Inference, and Blender Rendering. Benchmark scores directly influence your machine's pricing on the marketplace β higher scores command higher rental rates.
- β’Benchmark results are publicly visible on the FluxCore Leaderboard (leaderboard.fluxcore.ai)
- β’History view shows date, score, duration, and hardware used for each benchmark
- β’Compare your scores against network averages and top performers
- β’Run benchmarks periodically to keep your listing competitive
Auto-Switch: Mining β Rentals
The Auto-Switch feature maximizes your earnings by seamlessly transitioning between PoW mining and FluxEdge lease mode.
- β’When a rental starts, mining operations and overclocking (OC) profiles stop automatically
- β’When a rental ends, mining resumes with your default configuration and OC profile
- β’Ensures zero idle time between activities
- β’NiceHash integration is available as an additional mining fallback during idle periods
Provider Controls
| Feature | Description |
|---|---|
| Availability Windows | Set specific hours when your machine can be rented |
| Resource Allocation | Choose which GPUs and how much compute to offer |
| Pricing | Set competitive rates or use recommended marketplace pricing |
| Smart Scheduling | Automated availability based on usage patterns |
| Performance Monitoring | Track GPU temperature, utilization, power consumption |
| Bandwidth Control | Manage network usage for rental workloads |
| Multi-GPU Support | Offer multiple GPUs on a single machine |
| Emergency Stop | Instantly stop all rentals and reclaim full hardware control |
Set your availability windows to times when you are not using your machine. This maximizes rental income while keeping your GPU available for personal use during work hours.
Other articles in FluxEdge GPU Computing
What is FluxEdge?
Overview of the decentralized GPU compute marketplace β value proposition, network scale, and getting started.
Renting GPU Compute
How to rent Dedicated and Premium GPU machines β filtering, provisioning, and machine management.
Deploying Workloads
Quick Launch templates and custom Docker/YAML deployments β ports, domains, GPU selection, and persistent storage.
Pricing, Billing & Payments
Dynamic pricing formula, payment methods (fiat + crypto), deposit bonus, provider earnings, and KYC levels.
GPUs, Frameworks & Use Cases
Supported GPU models (RTX 4090 to H100), AI/ML frameworks, and real-world use cases from training to rendering.
Architecture, Security & Networking
Kubernetes orchestration, ArcaneOS chain-of-trust, container isolation, networking, and data encryption.
FluxEdge vs Traditional Cloud
Detailed comparison with AWS, GCP, Azure β egress fees, pricing, vendor lock-in, and migration strategy.