Score of 0
1 answer
130 views
Connect GPU to on-premise Kubernetes Cluster
Good morning, I would like to deploy AI applications in my Kubernetes cluster. Currently, my Kubernetes cluster has the following configuration: 3 master nodes 3 worker nodes I was thinking of ...
- reputation score 3
Score of 1
0 answers
95 views
Latency of warp add reduction instruction
The CUDA Programming Guide describe a warp instruction named __reduce_add_sync. What is the latency of the function, specifically in the Ampere architecture? Related sources: This table within the ...
- reputation score 576
Score of 4
1 answer
161 views
Can DRAM and SMEM instructions be issued in a single cycle?
In the Ampere architecture, consider the following scenarios: A single warp executes two load instructions: one from Shared Memory and one from DRAM. Two warps within the same SM, each executing a ...
- reputation score 576
Score of 0
1 answer
142 views
Arch linux with Nvidia driver
I decided to try arch linux and move all my development there. Currently I am using arch with sway. Now I am trying to setup my dev flow, but I am running into issues. Basically I have a project in C++...
- reputation score 127
Score of 0
1 answer
71 views
clone metadata in deepstream pipeline in python
I am currently implementing a deepstream pipeline in python with a tee split as follows: streammux -> tee -> queue1 -> detector1 -> tracker1 -> queue2 -> detector2 -> tracker2 The ...
- reputation score 1
Score of 0
2 answers
102 views
pycuda._driver.Error: cuInit failed: unknown error
I have a problem with pycuda. I used it for a python script i develop. I know this script work because i use it on other server. But on a specific server i got a problem : >>> import pycuda....
- reputation score 1