Glossary of Terms

• AI Inference: The process of using trained AI models to generate content, make predictions or decisions based on new data.
• GPU Compute: High-performance computing using GPUs for tasks like deep learning, data analytics, and rendering.
• Hosts: Data centers or hardware owners who list their GPUs on Nosana’s marketplace for rent.
• Clients: Businesses or individuals who rent GPU compute through Nosana Network for their workloads.
• Mainnet: The production-ready phase of Nosana Network, commonly used when addressing Web3 audiences.
• Product Launch: The term used for Web2 audiences to describe the launch of Nosana Network.
• Pay-Per-Second Billing: A flexible pricing model that charges clients based on the exact time they use GPU resources.
• Idle GPUs: GPU resources that are not being used and can be rented out through Nosana Network.
• Green Compute: Computing that prioritizes sustainability by optimizing idle resources and reducing e-waste.
• Reduced Carbon Footprint: Optimize idle resources to minimize environmental impact.
• GPU Marketplace: Nosana’s product where GPU compute resources are listed and rented, connecting Hosts and Clients for scalable and affordable computing.
• Consumer GPUs: GPUs designed for personal or gaming use
• Community market: A segment designed for new and unvalidated GPU hosts, offering them a place to start, test their setups, and prove reliability.
• Premium market: A segment designed for validated, high-performing GPU Providers offering reliable resources for critical workloads.
• Jobs: A compute workload run on Nosana network

---

Nosana’s Advantages

• Hassle-Free Management: Nosana delivers large quantities of GPUs without the complexities of infrastructure management.
• Secure and On-Demand Access: Tap into a global network of the latest GPUs, including hard-to-find models.
• Pay-Per-Second Billing: Flexible pricing that scales with your compute usage.
• Reduced Carbon Footprint: Optimize idle resources to minimize environmental impact.