What is rocJPEG?

AMD GPUs contain one or more media engines (VCNs). Each VCN has one or more JPEG engines that provide fully accelerated, hardware-based JPEG decoding. Hardware decoders consume lower power than CPU-based decoders. Dedicated hardware decoders offload decoding tasks from the CPU, boosting overall decoding throughput. With proper power management, decoding on hardware decoders can lower the overall system power consumption and improve decoding performance.

Using the rocJPEG API, you can decode compressed JPEG streams while keeping the resulting YUV images in video memory. With decoded images in video memory, you can run image post-processing using ROCm HIP, thereby avoiding unnecessary data copies via PCIe bus. You can post-process images using scaling or color space conversion and augmentation kernels (on a GPU or host) in a format for GPU/CPU-accelerated inferencing and training.

In addition, you can use the rocJPEG API to create multiple instances of JPEG decoder based on the number of available VCNs/JPEG engines in a GPU device. By configuring the decoder for a device, all available VCNs can be used seamlessly for decoding a batch of JPEG streams in parallel.