GPU Server Graphics Card Identification via BIOS Reset
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Graphics cards in GPU servers often fail to be identified after a restart, particularly affecting low-cost graphics cards like the AMD MI25, leading to impaired functionality.
Innovation Solution
A method involving obtaining correlation information during the BIOS startup process, determining if the graphics card belongs to a preset category, reading and resetting its memory address based on previous work cycle offsets, and sending a restart instruction to ensure proper identification and operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If low-cost graphics cards are applied in GPU servers, then purchasing cost is reduced, but the graphics cards fail to be identified after server restart
Solution Approach 1:
The patent applies preliminary action by performing a reset operation on the graphics card before the server restarts. This preliminary reset action ensures that the graphics card is in a proper state before the restart, preventing identification failures. The reset operation is executed in advance to prepare the graphics card for the upcoming restart cycle, thereby ensuring reliable identification without requiring expensive graphics cards.
2Reliability
If graphics card reset operation is added, then graphics card identification reliability is improved, but device complexity increases
Solution Approach 1:
The patent merges the reset operation with the existing graphics card identification process. Instead of adding a completely separate complex system, the reset function is integrated into the identification routine that already exists in the server. This merging approach allows the system to leverage existing code structures and hardware interfaces, thereby improving reliability while minimizing the increase in device complexity.
Data Source
AI summary
A method, a device, an apparatus for identifying a graphics card of a GPU server, and a medium are provided. The method includes: obtaining correlation information of a graphics card captured by an operation of enumerating PCI devices during a startup process of running a BIOS; determining whether the graphics card belongs to a preset category; reading a memory address of the graphics card in a configuration space of the PCI device in a case that the graphics card belongs to the preset category, and obtaining an actual memory address based on the memory address and an offset; and resetting the graphics card based on the actual memory address, and sending a restart instruction to perform a restart operation.