# GPU shares & Pricing
### What are GPU Shares?
GPU Shares are a **simple abstraction that allows users to participate in active AI model training** without managing hardware, infrastructure, or protocol integrations themselves.
Instead of renting GPUs or running nodes, users purchase **fractional output rights** to a live training node operated by AIyield.
Each GPU Share represents a **fixed fraction of the output** of a specific training node over a defined time period.
***
### The Node Abstraction
At the core of the system is the **AIyield Node**.
* **1 Node = 8× NVIDIA H100 GPUs**
* Nodes are allocated to **active training runs**
* Each run has a **fixed duration of 30 days**
* During this period, the node is continuously active on a training network
For the initial release, nodes are active on **NOUS Psyche Training**, a large-scale distributed model training environment.
***
### Fractionalization Into Shares
Each Node is **fractionalized into 400 equal GPU Shares per month**.
* **Shares per Node per 30 days:** 400
* **Price per Share:** $40
* **Duration:** 30 days (fixed)
This means:
* 1 GPU Share = **1/400 of the Node’s output** for that 30-day training run
* Users can purchase **multiple shares** to increase their proportional exposure
* Users can find their shares in the app Dashboard.
The abstraction is intentionally simple: shares represent **output participation**, not hardware ownership.
***
### Output Rights Formula
User output rights are calculated as:
```
User Output Rights
= (Shares Purchased / 400) × Node Output over 30 days
```
Examples:
* 1 share → 0.25% of node output
* 10 shares → 2.5% of node output
* 100 shares → 25% of node output
The total output depends on the training network, protocol rules, and actual training performance.
***
### What You Are (and Are Not) Buying
#### You are buying:
* A **time-limited right** to participate in the outputs of a specific training node
* Exposure to **real, active GPU usage** during a defined training run
* A fully managed experience (AIyield handles infra, orchestration, uptime, and protocol integration)
#### **Pricing & Payment**
GPU shares can be purchases through a monthly subscription, or on-chain through Solana or Base.\
Payment Methods
Users can pay **either with a card (subscription-style)** or **on-chain using crypto**. The underlying GPU Share abstraction is the same in both cases.
***
#### Card Payment (Subscription)
* Buy **1 GPU Share** as a monthly subscription
* Pay with a credit or debit card
* Fixed price: **$40 / month**
* Ideal for users who want:
* Simple onboarding
* Recurring participation
* embedded wallet setup. (Auto-rewards wallet setup through Privy)
**Notes**
* Subscription renews monthly unless cancelled
* Share participation is tied to the active Node run
* Auto-cancellation when the NOUS token drops.
***
#### Crypto Payment (On-Chain)
* Buy **1–450 GPU Shares** in a single transaction
* Pay with **USDC on Solana or Base**
* Price: **$40 per share**
* 10% Volume discounts may apply for larger purchases
* Rewards and outputs are **tracked on-chain**
**Notes**
* Shares are still time-bound (30 days)
* On-chain payment does not imply token ownership
* Accounting is handled transparently via the dashboard
***
#### Custom / High-Volume Plans
For users looking to deploy larger amounts of capital:
* Custom pricing and allocations
* Priority access to new Nodes
* Discounted rates
* Dedicated support
* Suitable for funds, DAOs, or power users
Access is handled manually via Discord.
**Disclaimer:**GPU Share Buyer Disclaimer
GPU Shares represent a **time-limited participation right** in the output of an AIyield Node during a defined training period. GPU Shares **do not constitute equity, debt, securities, derivatives, or ownership** in AIyield, the underlying hardware, or any associated protocol.
Purchasing a GPU Share grants the buyer a **proportional right to outputs generated by a specific Node** (8× NVIDIA H100 GPUs) over a **30-day training run**, calculated as:
> (Shares Purchased / 400) × Node Output over 30 days
#### Important Considerations
* **No guaranteed returns**: Outputs, rewards, or benefits are **not guaranteed** and may vary based on network conditions, protocol rules, model performance, or external factors.
* **Operational risk**: Training runs may be affected by hardware availability, network interruptions, protocol changes, or force-majeure events.
* **No secondary market**: GPU Shares are initially **non-transferable** and live off-chain unless explicitly stated otherwise.
* **Protocol dependency**: Outputs may depend on third-party networks (e.g. training protocols), which are **not controlled by AIyield**.
* **Not investment advice**: Participation should not be construed as financial or investment advice. Buyers are responsible for their own assessment of risks.
By purchasing a GPU Share, the buyer acknowledges and accepts these risks and confirms they understand that GPU Shares provide **access to output participation only**, not profit guarantees or ownership rights.
---
# Agent Instructions: Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter:
```
GET https://aiyield.gitbook.io/aiyield-docs/documentation/readme.md?ask=
```
The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.