Currently, TVM’s inference based on static models is single card inference. May I ask if TVM currently supports multi card distributed inference, or if this feature will be added in the future
Currently, TVM’s inference based on static models is single card inference. May I ask if TVM currently supports multi card distributed inference, or if this feature will be added in the future
Please see the related thread here
Okay, thank you. I will continue to monitor this research and development progress in the future@ Hzfengsy