GPU Workflow Allocation for Web-Based Drug Discovery Simulations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing drug candidate discovery systems face challenges in managing GPU resources and data consistency due to installation-based tools and lack of user-specific functionalities, making it difficult to integrate big data analysis and artificial intelligence effectively.
Innovation Solution
A multi-instance GPU-based system implemented as a web service platform with modules for project management, simulation setting, and workflow management, utilizing Docker containers and virtual GPUs to allocate resources efficiently and manage complex simulation workflows.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If installation-based data analysis tools are used, then data processing functionality is provided, but accessibility and ease of operation deteriorate
Solution Approach 1:
The patent replaces the mechanical installation-based tool execution model with a web service-based system. Users access drug discovery simulation tools through web browsers without installing software, substituting the traditional installation-execution mechanism with network-based service access. This resolves the contradiction by eliminating installation requirements while maintaining full data processing functionality.
Solution Approach 2:
The web service platform provides a universal access interface that consolidates multiple drug discovery tools and functions into a single system. Instead of requiring separate installations for different analysis tools, the platform offers integrated project management, simulation settings, workflow management, and execution capabilities through a unified web interface, improving accessibility while maintaining comprehensive functionality.
2Adaptability or versatility
If multiple separate tools are installed for big data analysis, then analytical functionality is improved, but data sharing and consistency deteriorate
Solution Approach 1:
The patent merges multiple separate analytical tools into a single integrated web service platform. The system combines project management, simulation configuration, workflow design, and execution capabilities into one unified environment. This integration ensures that all tools access and share the same data sources and results consistently, eliminating data fragmentation while preserving the full analytical functionality of individual tools.
Solution Approach 2:
The web service platform acts as an intermediary layer between users and multiple analytical tools. Instead of directly installing and managing separate tools, users interact with the platform which mediates access to underlying computational resources and tools. This intermediary architecture ensures consistent data management and sharing across all analytical functions while maintaining versatility.
3Adaptability or versatility
If different operating systems are used for tool installation, then system compatibility is improved, but data management consistency deteriorates
Solution Approach 1:
The patent replaces the operating system-dependent tool installation model with a web-based service architecture. Instead of installing tools that must be configured for each operating system, users access the drug discovery platform through web browsers on any operating system. This substitution eliminates OS-specific configuration issues while maintaining universal accessibility and consistent data management across different platforms.
4Productivity
If GPU resources are allocated to individual tasks, then task execution efficiency is improved, but resource utilization and flexibility deteriorate
Solution Approach 1:
The patent implements dynamic GPU resource allocation within the web service platform. Instead of static one-to-one task-GPU mapping, the system dynamically assigns GPU resources to tasks based on current workload, task requirements, and resource availability. This dynamic allocation mechanism maintains high execution efficiency while providing flexible adaptability to different task types and resource conditions.
Solution Approach 2:
The patent segments GPU resources into multiple virtual GPU instances that can be independently allocated to different tasks. This segmentation allows the system to divide physical GPU resources into smaller allocable units, enabling flexible resource distribution across multiple concurrent tasks while maintaining efficient utilization of the underlying hardware resources.
Data Source
AI summary
A drug candidate discovery system may include: a project management module configured to create a project for adding a task to perform drug candidate discovery; a simulation management module configured to create a simulation on the created project; a simulation setting module configured to set a simulation workflow for the simulation based on input from a user, using a canvas area and a simulation setting area containing a protein structure data input area and a task module selection area, each containing one or more objects that can be dragged and dropped onto the canvas area and converted into a node; a simulation workflow management module configured to manage information on nodes that can precede or follow in the simulation workflow; and a simulation execution module configured to manage and execute a task for running the simulation workflow by dividing and allocating GPU resource to the respective nodes.