Executable Graph Reuse Through Dynamic Task Graph Modification

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing executable graphs are typically usable for only a single workload of a task graph from which they were created, limiting their versatility in performing workloads associated with new task graphs.

Innovation Solution

Techniques for modifying an executable graph to adapt it for performing workloads associated with new task graphs, allowing reuse across multiple task graphs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If an executable graph is created from a task graph, then the workload can be performed by configuring computing resources according to the executable graph, but the executable graph is typically usable for only a single workload and cannot be reused for new task graphs

Engineering Contradiction:

Improvereusability of executable graphVSAvoidconfiguration complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The executable graph is designed to perform multiple functions by enabling it to handle workloads from different task graphs. The system allows the executable graph to be configured for various computing resources and reused across multiple task graphs, making a single executable graph serve universal purposes rather than being dedicated to a single workload.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The executable graph incorporates dynamic configuration capabilities that allow it to adapt its structure and parameters based on the specific task graph being executed. This dynamic nature enables the same executable graph to be flexibly reconfigured for different workloads while maintaining optimized performance characteristics.

Inventive Principle:

Principle #15Dynamics

2Productivity

If an executable graph is created specifically for a single task graph workload, then optimization for that specific workload is achieved, but the executable graph cannot be reused for other task graphs, increasing configuration overhead

Engineering Contradiction:

Improveworkload execution efficiencyVSAvoidconfiguration time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The executable graph is prepared in advance with pre-configured computing resources and optimized structures that can be quickly adapted to different task graphs. By having the executable graph ready with preliminary configurations, the system reduces the time required to set up new workloads while maintaining execution efficiency.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system enables rapid reconfiguration of the executable graph by changing key parameters such as computing resource allocations, data transfer configurations, and task scheduling parameters. This allows the executable graph to maintain high productivity across different workloads while minimizing configuration time through parameter adjustment rather than complete reconfiguration.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20260079755A1Techniques for modifying an executable graph to perform a workload associated with a new task graph

Publication Date: 2026.03.19 NVIDIA CORP

AI summary

Techniques to modify executable graphs to perform different workloads. In at least one embodiment, an executable version of a first task graph is modified by applying a non-executable version of a second task graph to executable version of first task graph so that executable version of first task graph can perform a second workload of non-executable version of second task graph.