Portable Framework for Heterogeneous Compute Systems

The patent introduces a portable framework as an intermediary layer between the software application and heterogeneous hardware resources. This framework includes a task dispatcher that translates high-level task descriptions into hardware-specific operations, and a resource manager that abstracts hardware details. The intermediate representation (IR) format serves as a mediator that can be optimized for different hardware platforms without requiring static binding, thus achieving hardware optimization while maintaining software portability and reducing development complexity.

The patent creates a universal portable framework that can execute on multiple heterogeneous hardware platforms including CPUs, GPUs, FPGAs, and other accelerators. The framework uses a hardware-agnostic intermediate representation that can be deployed across different infrastructure environments without modification. This multi-functionality approach allows the same software to adapt to various hardware configurations, achieving both high performance on specific hardware and broad portability across platforms, while reducing infrastructure deployment costs through shared codebase.

The patent implements dynamic task dispatching that can adaptively allocate computational tasks to appropriate hardware resources based on runtime conditions, task characteristics, and hardware availability. The portable framework maintains a pool of heterogeneous computing resources and dynamically binds tasks to optimal execution platforms without static hardware binding. This dynamic approach ensures high execution efficiency by matching task requirements with suitable hardware capabilities while maximizing resource utilization and preventing hardware underutilization across the infrastructure.