2 Broken H100 GPU's
I'm 14, paid for 2x H100 GPUs for my AI research company Aedis, and got completely broken hardware.
ERROR: "CUDA out of memory. GPU 0 has 79.18 GiB total, 14.50 MiB free. 79.16 GiB already in use"
How is 79GB already consumed on a FRESH session?? My 2B model can't even load.
My code works perfectly on CPU, proving your GPUs are hardware-faulty.
Support says "wait 1-2 business days" while I'm being charged NOW for broken equipment. I only have $2 left.
Screenshots available. Ticket #22071.
ERROR: "CUDA out of memory. GPU 0 has 79.18 GiB total, 14.50 MiB free. 79.16 GiB already in use"
How is 79GB already consumed on a FRESH session?? My 2B model can't even load.
My code works perfectly on CPU, proving your GPUs are hardware-faulty.
Support says "wait 1-2 business days" while I'm being charged NOW for broken equipment. I only have $2 left.
Screenshots available. Ticket #22071.
