ComfyUI Node: 🔥 Dynamic RAM Cache Control

🔥 Dynamic RAM Cache Control:

The DynamicRAMCacheControl node is designed to manage memory usage efficiently within the ComfyUI environment by dynamically switching between different caching modes. Its primary purpose is to optimize RAM usage by either maintaining a classic cache mode, which does not evict data, or a RAM pressure mode, which automatically purges data to maintain a specified amount of free RAM. This node is particularly beneficial for workflows that require adaptive memory management, ensuring that your system remains responsive and efficient even under varying workloads. By intelligently managing cache data, it helps prevent out-of-memory errors and enhances the overall performance of your AI art generation tasks.

🔥 Dynamic RAM Cache Control Input Parameters:

mode

The mode parameter determines the caching strategy employed by the node. It offers two options: "CLASSIC (No Eviction)" and "RAM_PRESSURE (Auto Purge)". The "CLASSIC" mode retains all cached data without eviction, which is suitable for environments with ample memory. In contrast, the "RAM_PRESSURE" mode automatically purges less critical data to maintain a specified level of free RAM, making it ideal for systems with limited memory resources. The default setting is "RAM_PRESSURE (Auto Purge)", allowing for automatic memory management.

cleanup_threshold

The cleanup_threshold parameter specifies the minimum amount of free RAM, in gigabytes, that the system should maintain when operating in RAM pressure mode. This threshold helps determine when the cache should start purging data to free up memory. The parameter accepts a floating-point value ranging from 0.1 to 256.0, with a default value of 2.0 GB. Adjusting this threshold allows you to fine-tune the balance between memory usage and performance, ensuring that your system remains stable and efficient.

🔥 Dynamic RAM Cache Control Output Parameters:

output_passthrough

The output_passthrough parameter serves as a conduit for any input data passed to the node. It ensures that the input data is returned unchanged, allowing for seamless integration with other nodes in your workflow. This output is particularly useful when you want to maintain the flow of data through your node graph without interruption, even as the node manages memory in the background.

🔥 Dynamic RAM Cache Control Usage Tips:

• To optimize performance on systems with limited RAM, use the "RAM_PRESSURE (Auto Purge)" mode and set an appropriate cleanup_threshold to ensure sufficient free memory is maintained.
• For workflows that require consistent data retention and have ample memory resources, consider using the "CLASSIC (No Eviction)" mode to prevent unnecessary data purging.

🔥 Dynamic RAM Cache Control Common Errors and Solutions:

[DynamicRAMCache] RAMPressureCache class not available in caching module

• Explanation: This error occurs when the RAMPressureCache class is not found in the caching module, which may happen if you are using an outdated version of ComfyUI.
• Solution: Ensure that you are using a ComfyUI version released after October 31, 2025, or switch to the "CLASSIC" mode if the RAMPressureCache class is unavailable.

[DynamicRAMCache] PromptExecutor not found.

• Explanation: This warning indicates that the PromptExecutor object, which is necessary for cache management, could not be located.
• Solution: Verify that your workflow includes a PromptExecutor and that it is correctly configured to interact with the DynamicRAMCacheControl node.

[DynamicRAMCache] CacheSet has no 'outputs' attribute.

• Explanation: This warning suggests that the CacheSet object does not have the required outputs attribute, which is essential for cache operations.
• Solution: Check the configuration of your CacheSet object to ensure it includes the outputs attribute, or consult the documentation for proper setup instructions.