Question
SYS-2027GR-TRF: Has 4 x Nvidia Tesla K80-- GPU #4 in Perf mode P8 (but all other GPUs in P0) and has double-bit ECC unrecoverable errors and power and memory usage much lower than other GPUs - Random application crashes. Segmentation faults. Only happens on this server, other nodes run fine.
Print
Answer
Double-bit ECC unrecoverable errors likely due to bad GPU. You may try to swap the GPU#4 that has double bit error with another working GPU to see if the problem follows the GPU#4 or not.
Was this FAQ helpful?
Enter Comments Below:
Note: Your comments/feedback should be limited to this FAQ only. For technical support, please send an email to .
Enter your email address below if you'd like technical support staff to reply:
Please type the Captcha (no space)
Y 4 X P
FAQ ID:
24617
Related Category / Keyword:
N/A
Date Posted:
12/16/16
Code:
Find more questions
Still Need Support?
Having trouble finding what you're searching for? Contact our support team for further assistance.

Certain products may not be available in your region