■ This is an article posted on June 2018, 2, so the content of the information may be out of date.
A customer asked us about a workstation for Deep Learning. As a result of hearing about the usage environment, it has excellent half-precision calculation and video memory capacity.TESLA P100We made a proposal with the installed machine.
【Main Specifications】
| CPU | Xeon E5-2680v4 (2.4GHz 14 cores) x2 |
| memory | 128GB ECC REG |
| storage | System: SSD 960GB Data: HDD 2TB |
| network | GigabitLAN x2 |
| video | TESLA P100 16GB + GPU Kit |
| Housing + power supply | 4U tower chassis + 2000W redundant power supply (100W at 1000V) |
| OS | Ubuntu 16.04 64bit |
| Others | CUDA Toolkit installation |
As a precaution, due to the problem of the cooling mechanism, it is basically impossible to mix P100 and a video card such as GeForce. Also, Tesla P100 installed in this machine cannot be logged in with the GUI in combination with Ubuntu 16.04. If you need to work in the GUI environment, you need a separate video card.
| ■ Click here for details and inquiries about this PC case WS example for deep learning with TESLA P100 |
