Method of operating shared GPU resource and a shared GPU device
AssigneeHuawei Cloud Computing Technologies Co., Ltd.
InventorsOmer Anson, Dima Kuznetsov
Life of this patent
Estimating the term…
Abstract
A method of operating a shared GPU device includes receiving GPU software from each of one or more tenants sharing the GPU device. The method further includes for each received software, verifying the software is safe to execute, and compiling each software that is verified as safe. The method further includes executing each compiled software in turn using the shared GPU device.
Claims20
A method implemented by a graphics processing unit (GPU) device, the method comprising:
- receiving GPU softwares from each of one or more tenants sharing the GPU device;
- verifying that the GPU softwares are safe to execute by:
- checking one or more entry code blocks of the GPU softwares to determine one or more memory addresses accessed by the one or more entry code blocks;
- and
- verifying that the GPU softwares are safe when no unsafe memory address is accessed;
- compiling, when verified as safe, the GPU softwares to obtain compiled GPU softwares;
- and
- executing the compiled GPU softwares using the shared GPU device.
A graphics processing unit (GPU) device comprising:
- a memory configured to store instructions;
- and
- one or more processors coupled to the memory and configured to execute the instructions to cause the GPU device to:
- receive GPU softwares from each of one or more tenants sharing the GPU device;
- verify that the GPU softwares are safe to execute by:
- checking one or more entry code blocks of the GPU softwares to determine one or more memory addresses accessed by the one or more entry code blocks;
- and
- verifying that the GPU softwares are safe when no unsafe memory address is accessed;
- compile, when verified as safe, the GPU softwares to obtain compiled GPU softwares;
- and
- execute the compiled GPU softwares using the shared GPU device.
A computer program comprising computer-executable instructions that are stored on a non-transitory machine-readable storage medium and that, when executed by one or more processors, cause an electronic device to:
- receive graphics processing unit (GPU) softwares from each of one or more tenants sharing a GPU device;
- verify that the GPU softwares are safe to execute by:
- checking one or more entry code blocks of the GPU softwares to determine one or more memory addresses accessed by the one or more entry code blocks;
- and
- verifying that the GPU softwares are safe when no unsafe memory address is accessed;
- compile, when verified as safe, the GPU softwares to obtain compiled GPU softwares;
- and
- execute the compiled GPU softwares using the shared GPU device.
Dependencies are inferred from the claim text.