Find high-impact opportunities to offload to a target device and the profitability of porting your code to a target graphics processing unit (GPU) by running the Offload Modeling perspective.
The Offload Modeling perspective can help you to:
Determine if you should offload your code to a target device and what is the potential speedup for different GPU platforms before getting a hardware
Identify loops that are recommended for offloading
Pinpoint potential performance bottlenecks on the target platform to decide on optimization directions before offloading the code
Check how effectively data can be transferred between baseline and target devices after you offload your code
With the Offload Modeling perspective, the following workflows are available:
- CPU-to-GPU offload modeling for C, C++, and Fortran applications: Analyze a C, C++, or Fortran application and model its performance on a target GPU platform. Use this workflow to find offload opportunities and prepare your code for efficient offload to the GPU.
- CPU-to-GPU offload modeling for Data Parallel C++ (DPC++), OpenMP* target, and OpenCL™ applications: Analyze a DPC++, OpenMP target, or OpenCL application offloaded to a CPU and model its performance on a target GPU platform. Use this workflow to understand how you can improve performance of your application on the target GPU and check if your code has other offload opportunities. This workflow analyzes both CPU and GPU parts of your application.
- GPU-to-GPU offload modeling for DPC++, OpenMP target, and OpenCL applications (technical preview): Analyze DPC++, OpenMP target, or OpenCL application running on a GPU and model ita performance on a different GPU platform. Use this workflow to understand how you can improve your application performance and check if you can get a higher speedup if you offload the application to a different GPU platform.
Note
Currently, you can model application performance only on Intel® GPUs.How It Works
The Offload Modeling perspective includes the following steps:
- Get the baseline performance data for your application by running a Survey analysis.
- Identify the number of times loops are invoked and executed and the number of floating-point and integer operations, estimate cache and memory traffics on target device memory subsystem by running the Characterization analysis.
- Mark up loops of interest and identify loop-carried dependencies that might block parallel execution by running the Dependencies analysis.
- Estimate the total program speedup on a target device and other performance metrics according to Amdahl's law, considering speedup from the most profitable regions by running Performance Modeling. A region is profitable if its execution time on the target is less than on a host.
Offload Modeling Summary
Offload Modeling perspective measures performance of your application and compares it with its modeled performance on a selected target GPU so that you can decide what parts of your application you can execute on the GPU and how you can optimize it to get a better performance after offloading.
Main metrics for the modeled performance of your program indicating if you should offload your application to a target device or not.
Specific factors that prevent your code from achieving a better performance if executed on a target device, that is the factors that your code is bounded by.
Top five offloaded loops/functions that provide the highest benefit and top five non-offloaded loops/functions with why-not-offloaded reasons.
