我们有一个应用程序，可以在金属上运行一些相当繁重的机器学习代码。最近，我们发现gpu挂起错误有所增加，这似乎与正在运行的系统动画（目前为UIView.animate）一致，这使我相信这与GPU上的过多负载有关。我们的机器学习代码在侧线程上运行，因为它还具有一些CPU组件，而且非常密集，需要在主线程上运行afaik UI动画。想知道是否有人曾经遇到过此类问题，以及是否有任何好的解决方法？
2018-12-11 11:38:41.203756-0800 Onyx[46615:7879990] Execution of the command buffer was aborted due to an error during execution. Caused GPU Timeout Error (IOAF code 2)
2018-12-11 11:38:41.203895-0800 Onyx[46615:7879990] Execution of the command buffer was aborted due to an error during execution. Caused GPU Timeout Error (IOAF code 2)
2018-12-11 11:38:41.254021-0800 Onyx[46615:7879990] Execution of the command buffer was aborted due to an error during execution. Caused GPU Hang Error (IOAF code 3)
2018-12-11 11:38:41.254058-0800 Onyx[46615:7879990] Execution of the command buffer was aborted due to an error during execution. Caused GPU Hang Error (IOAF code 3)
2018-12-11 11:38:41.254090-0800 Onyx[46615:7879990] Execution of the command buffer was aborted due to an error during execution. Caused GPU Hang Error (IOAF code 3)
2018-12-11 11:38:41.254100-0800 Onyx[46615:7879990] Execution of the command buffer was aborted due to an error during execution. Caused GPU Hang Error (IOAF code 3)