A6000-equipped exercise machine

We received a consultation from a customer about a dedicated machine for image analysis and machine learning in a class at a university.
Assuming that up to 30 students will use it, they will access it from their own PCs and practice scientific calculations, image analysis, and machine learning.

The OS is assumed to be Ubuntu, and we would like to ship it with Anaconda / SSH-Server / R / TensorFlow / Pytorch installed.
You are also considering using GPU, but since it is used infrequently, there is no problem in lowering the specifications depending on the cost, so it is a condition that we want to keep the budget of about 400 million yen as a whole.

Since it is difficult to fit in the budget when considering with full-scale GPU server specifications, the proposal is for a GPU x 1 sheet / 100V power supply environment.
To balance the budget, the CPU is 64 cores (32 cores x 2) and the memory is 256GB. When used by 30 people, the resource allocation is about CPU 2 cores (4 threads) / memory 8GB per person.

The scope of work is assumed to be until the specified software is installed.If you need individual settings, please feel free to contact us.

A paid license is required to use Anaconda for commercial activities.
Until now, we have only handled Enterprise for large-scale use, but now we can handle licenses for small-scale use.For details, please see below.

【Main Specifications】

CPU EPYC MILAN 7543 (2.80GHz 32 cores) x2
memory 256GB REG ECC
Storage 1 2TB SSD S-ATA
Storage 2 4TB M.2 SSD
video Nvidia RTX A6000
network on board (10GBase-T x2)
Housing + power supply Tower type housing + 1600W
OS Ubuntu 20.04
Others CUDA Toolkit 11
Anaconda/SSH-Server/R/TensorFlow/Pytorch