Currently, there is no official template or way to Monitor GPUs in servers. GPUs are a necessary part of next-generation AI-based applications that are using accelerated computing.
I found below 2 templates to monitor GPUs via Zabbix that are really old and have not been updated since 2018. Although working with Zabbix 5.0 but we need more data and deep monitoring for GPUs.
1. Zabbix Share - NVidia-SMI Integration