Abstract: A GPU simulation method. An instruction sequence of a client GPU is intercepted in a kernel state simulator based on system virtualization and GPU using principle, and a mechanism is selected according to user configuration to accomplish simulation of the client GPU. In first mechanism, instruction translation is accomplished on low-level semantics based on a binary translation technology, and instructions are executed on a host GPU; in second mechanism, instruction conversion is accomplished using an existing GPU software stack, and instructions are executed on host GPU. The method provides an efficient simulated GPU for a virtual machine based on a host machine physical GPU, and solves the problem of slow GPU simulation. Based on a system virtualization technology and by virtue of a convenient condition provided by an existing GPU software stack, the GPU simulation speed is improved, and the implementation difficulty and complexity of the method are effectively controlled.

Abstract: A memory sharing method of virtual machines through the combination of KSM and pass-through, including: a virtual machine manager judging whether operating systems of guests use IOMMU, if not, not participating in shared mapping of a KSM technology; if yes, judging memory pages of each guest to confirm whether the pages are mapping pages, if yes, remain the mapping pages into a host; and if not, on the premise of keeping the properties of Pass-through, using the KSM technology for all non-mapping pages to merge the memory pages with same contents among various virtual machines and perform write protection processing simultaneously. The guest memory pages are divided into those special for DMA and those for non-DMA purpose, then the KSM technology is only selectively applied to the non-DMA pages, and on the premise of keeping the properties of Pass-through, the object of saving memory resources is achieved simultaneously.