Abstract: Obtain, at a computing device, a segment of computer code. With a classification module of a machine learning system executing on the computing device, determine a required annotation category for the segment of computer code. With an annotation generation module of the machine learning system executing on the computing device, generate a natural language annotation of the segment of computer code based on the segment of computer code and the required annotation category. Provide the natural language annotation to a user interface for display adjacent the segment of computer code.

Abstract: An image processing method and a computer system. The method may be applied to a cloud-side server in a cloud mobile phone. The server may be a virtualization server, a host operating system and a guest operating system are deployed on the server, a user mode graphics driver is deployed in the guest operating system, and a kernel mode graphics driver is deployed in the host operating system. The user mode graphics driver and the kernel mode graphics driver collaborate with each other to implement image rendering of the server. Then, the server may send a rendered image to the cloud mobile phone. Accordingly, an instruction translation process is reduced, to reduce overheads of a processor and improve image processing efficiency.

Abstract: An image processing method and a computer system. The method may be applied to a cloud-side server in a cloud mobile phone. The server may be a virtualization server, a host operating system and a guest operating system are deployed on the server, a user mode graphics driver is deployed in the guest operating system, and a kernel mode graphics driver is deployed in the host operating system. The user mode graphics driver and the kernel mode graphics driver collaborate with each other to implement image rendering of the server. Then, the server may send a rendered image to the cloud mobile phone. Accordingly, an instruction translation process is reduced, to reduce overheads of a processor and improve image processing efficiency.

Abstract: Embodiments of the present application provide a method for implementing a virtual GPU. The method for implementing a virtual GPU includes: allocating to each of the virtual GPUs a running time slice corresponding to the resource requirement of the virtual GPU according to resource requirements of virtual GPUs running on the same physical GPU, wherein a sum of running time slices of all virtual GPUs configured on a physical GPU is less than or equal to a scheduling period; and allocating resources of the physical GPU to the virtual GPUs according to the running time slices allocated to the virtual GPUs.

Abstract: A request is sent from a new virtual function (VF) to a physical function for requesting the initialization of the new VF. The controlling physical function and the new VF establish a two-way communication channel that to start and end the VF's exclusive accesses to registers in a configuration space. The physical function uses a timing control to monitor that exclusive register access by the new VF is completed within a predetermined time period. The new VF is only granted a predetermined time period of exclusive access to complete its initialization process. If the exclusive access period is timed out, the controlling physical function can terminate the VF to prevent GPU stalls.

Abstract: Embodiments of the present application provide a method for implementing a virtual GPU. The method for implementing a virtual GPU includes: allocating to each of the virtual GPUs a running time slice corresponding to the resource requirement of the virtual GPU according to resource requirements of virtual GPUs running on the same physical GPU, wherein a sum of running time slices of all virtual GPUs configured on a physical GPU is less than or equal to a scheduling period; and allocating resources of the physical GPU to the virtual GPUs according to the running time slices allocated to the virtual GPUs.

Abstract: A request is sent from a new virtual function (VF) to a physical function for requesting the initialization of the new VF. The controlling physical function and the new VF establish a two-way communication channel that to start and end the VF's exclusive accesses to registers in a configuration space. The physical function uses a timing control to monitor that exclusive register access by the new VF is completed within a predetermined time period. The new VF is only granted a predetermined time period of exclusive access to complete its initialization process. If the exclusive access period is timed out, the controlling physical function can terminate the VF to prevent GPU stalls.

Abstract: A time stamp value associated with a virtual function of a guest virtual machine (VM) is periodically updated. One of a plurality of idle worker threads in a thread pool is assigned to periodically increment the time stamp value after initialization of an instance of the guest VM. An inactive status of the guest VM is detected based at least in part on the time stamp value not changing over a specified time period. The provision of resources to the virtual function of the inactive guest VM can be terminated based on its inactive status. In one embodiment, the virtual function is associated with a graphics processing unit (GPU) and terminating the provision of resources includes terminating the scheduling of cycles of GPU time.