Complex rendering using tile buffers

A complex rendering process comprising:

• performing a rendering process of the complex rendering process in which one or more tile buffers are configured to store rendered fragment data relating to a tile;
• and performing a machine learning process of the complex rendering process in which the one or more tile buffers are configured to store machine learning data relating to the machine learning process,
• wherein
• generating a job list;
• determining whether at least part of a result of a task to perform the complex rendering process will be required by an other shader core, and
• in response to determining that the at least part of the result of the complex rendering process will be required by the another shader core, causing the generation of task affinity information for the task.

A non-transitory computer-readable medium to store computer-readable code for fabrication of an apparatus configured to operate as a tile-based graphics processing system comprising:

• a shader core configured to perform a complex rendering process comprising a rendering process and a machine learning process, the shader core comprising:
• one or more tile buffers configured to store data locally to the shader core;
• wherein
• during the rendering process of the complex rendering process, the one or more tile buffers are configured to store rendered fragment data relating to a tile;
• and
• during the machine learning process of the complex rendering process, the one or more tile buffers are configured to store machine learning data relating to the machine learning process
• wherein
• the tile-based graphics processing system comprises a driver configured to generate a job list;
• the driver is configured to determine whether at least part of a result of a task to perform the complex rendering process will be required by an other shader core;
• and
• in response to the determination being that the at least part of the result of the complex rendering process will be required by the another shader core, the driver is configured to cause the generation of task affinity information for the task.

A tile-based graphics processing system comprising:

• a shader core to perform a complex rendering process comprising a rendering process and a machine learning process, the shader core comprising:
• one or more tile buffers configured to store data locally to the shader core;
• wherein
• during the rendering process of the complex rendering process, the one or more tile buffers are configured to store rendered fragment data relating to a tile;
• and
• during the machine learning process of the complex rendering process, the one or more tile buffers are configured to store machine learning data relating to the machine learning process;
• the feature map corresponds to the tile;
• and
• the machine learning process is based on rendered fragment data corresponding to the tile and one or more neighbouring tiles of the tile;
• the apparatus is configured to determine whether at least part of a result of the complex rendering process will be required by an other apparatus;
• and
• in response to the determination being that the at least part of the result of the complex rendering process will be required by the other apparatus, the apparatus is configured to cause the at least part of the result of the complex rendering process to be preferentially stored in a cache.

A tile-based graphics processing system comprising:

• a shader core to perform a complex rendering process comprising a rendering process and a machine learning process, the shader core comprising:
• one or more tile buffers configured to store data locally to the shader core;
• wherein
• during the rendering process of the complex rendering process, the one or more tile buffers are configured to store rendered fragment data relating to a tile;
• and
• during the machine learning process of the complex rendering process, the one or more tile buffers are configured to store machine learning data relating to the machine learning process
• comparison circuitry configured to compare a hash of the feature map with a hash of a result of performing the complex rendering process data and to store the result back to a storage circuit in dependence on the hash of the feature map and the hash of the result differing.