I just got nvidia-docker up on my wife's intro to deep learning rig and have been running monero mining just to put the GPU under some load and ensure it is stable like that before she uses it. I decided to make her a smaller machine while she is in her early learning period with CUDA, cudnn, Tensorflow and others.
Here is the configuration:
Here is the configuration:
- Supermciro X10SDV-TLN4F (Xeon D-1540)
- ASUS Dual Fan NVIDIA GTX 1060 6GB
- 64GB DDR4-2133
- 4x 480GB Intel DC S3500 RAID 10
- Icy Dock 4-in-1 hot swap enclosure
- All 4 NICs lit up but just passing mining traffic
- One additional 120mm Corsair PWM fan
- Seasonic fanless Platinum 520w PSU
- Ubuntu 16.04.1
- Monero pegs the GPU but RAM wise uses 733MiB / 6104MiB
- CPU utilization otherwise is around 1%
- System memory is using around 500MB/ 62GB