海光k100-AI安装驱动报错,驱动版本为rock-6.3.28。
共八张卡,且在BMC界面查看发现掉了一张卡,是不是说明硬件有问题啊。
hycu: Failed to send message 0x4, Parameter 0xf, response 0xfe
hycu: PowerPlay hw init failed
hycu: hw init of IP block PowerPlay failed -5
hycu: hycu\_device\_ip\_init failed
hycu: Fatal error during GPU init
HW online 5 ok, offline 3
offline: [0000:45:00.0, 0000:48:00.0, 0000:30:00.0]