-
Hi Guys
I have built a new install on a notebook with a K610M GPU. Running CentOS7.4.
01:00.0 VGA compatible controller: NVIDIA Corporation GK208GLM [Quadro K610M] (rev a1)
No errors were seen…
-
Please make sure that this is a build/installation issue. As per our [GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md), we only address code/doc bugs, performance issues,…
-
Like in XenCenter:
![](https://pbs.twimg.com/media/DHvnpENXsAAfRgp.jpg)
- [ ] detect if there is GPU counters in the RRDs
- [ ] if any, add the GPU graph
-
Sometimes GPU perf state change to P8 and cause low performance for running job. Need to investigate why it happens.
If it's an abnormal issue, need to add alert.
-
**System information**
- Have I written custom code (as opposed to using a stock example script provided in TensorFlow): Yes
- OS Platform and Distribution (e.g., Linux Ubuntu 16.04):
- TensorFlow …
-
I install this git follow the instruction
https://www.yuque.com/mnn/en/build_linux
It success and I got this file libMNN.so but I can not import the MNN from python. Anyone can help me ?
-
I'm running a simulation with ~300,000 atoms so I bench-marked the speed across 3 NVDIA GTX Titan GPUs. Here are the results.
https://github.com/choderalab/gpcr-simulation/blob/master/beta2_2rh1/Wei…
-
On my Openshift 4.3 I have two worker nodes with GeForce cards:
```
[root@worker01 ~]# lspci|grep -i nvidia
06:00.0 VGA compatible controller: NVIDIA Corporation GP107 [GeForce GTX 1050 Ti] (rev a1…
-
## Expected behavior
i'm following Basic workflow video and data_dst and data_src extract faces (any bat) don't work
## Actual behavior
all images found
0 faces detected
## Steps to repr…
-
Hi, there,
From this page:
https://github.com/microsoft/Azure-Kinect-Sensor-SDK/blob/develop/docs/building.md
It is mentioned that we shall use **VS2017** in order to build and install the …