[tsibel]

Hello everyone


I need your help for my configuration. We will buy a workstation for our CFD group at the university. We will model mostly two phase flow

CPU 2xIntel Xeon Gold X2 6140 or 2xIntel Xeon Gold X2 6148
RAM: 64GB LRDIMM Samsung DDR4-2666, CL19, reg. ECC (8x 64GB = 512GB) or 8X32=256GB
NVIDIA Quadro P2000 5 GB GDDR5
Mainboard: ASUS WS C621E Sage, Dual So. 3647; E-ATX
SSD: 512GB Samsung 970 Pro, M.2 PCIe (MZ-V7P512BW)
HDD: 6TB Seagate IronWolf Pro NAS, SATA3, 7200RPM (ST6000NE0023)
And some questions:
1) Do you have a suggestion for the cooling system?
2) Does the GPU provide additional performance for academic studies? Or should the money for CPU?
3) I've searched the AVX512, but I don't really understand. It seems that it has both negatives and positives?


Thanks

[flotus1]

Edit: moved my answer to the new thread

[ztdep]

Originally Posted by tsibel

Hello everyone


I need your help for my configuration. We will buy a workstation for our CFD group at the university. We will model mostly two phase flow

CPU 2xIntel Xeon Gold X2 6140 or 2xIntel Xeon Gold X2 6148
RAM: 64GB LRDIMM Samsung DDR4-2666, CL19, reg. ECC (8x 64GB = 512GB) or 8X32=256GB
NVIDIA Quadro P2000 5 GB GDDR5
Mainboard: ASUS WS C621E Sage, Dual So. 3647; E-ATX
SSD: 512GB Samsung 970 Pro, M.2 PCIe (MZ-V7P512BW)
HDD: 6TB Seagate IronWolf Pro NAS, SATA3, 7200RPM (ST6000NE0023)
And some questions:
1) Do you have a suggestion for the cooling system?
2) Does the GPU provide additional performance for academic studies? Or should the money for CPU?
3) I've searched the AVX512, but I don't really understand. It seems that it has both negatives and positives?


Thanks

GPU with 5G memory do nothing better, the data movement between the local memory and GPU memory through PCIe is very time-consuming.
For a workstation, we will benefit more from a CPU with many cores.
Palatalization efficiency strongly depends on your algorithm. Spend more time to optimize it.

[jmex]

Originally Posted by ztdep

I strongly suggest you buy the old Xeon workstations. For example, I now use a workstation with two Xeon 2650 v2 CPUs (2.60GHz, 8 cores, 16threads) 32G RAM. It is very cheap now. You can buy two workstations to set up a small cluster. Now, we have totally 32 Cores and 64 threads to conduct the simulation.
Another important suggestion is your operating system. fluent in the Linux system run obviously more quickly than in the windows system.
This is my experience, hope this will help you.

How did you merge 8 cores to form 32 cores?

[flotus1]

8 cores each in a dual-socket system already gets you 16 cores with shared memory. Then you can use some kind of interconnect (Ethernet, Infiniband...) to form a distributed memory cluster with as many cores as you can afford.

[amirkh]

Originally Posted by ztdep

I strongly suggest you buy the old Xeon workstations. For example, I now use a workstation with two Xeon 2650 v2 CPUs (2.60GHz, 8 cores, 16threads) 32G RAM. It is very cheap now. You can buy two workstations to set up a small cluster. Now, we have totally 32 Cores and 64 threads to conduct the simulation.
Another important suggestion is your operating system. fluent in the Linux system run obviously more quickly than in the windows system.
This is my experience, hope this will help you.

*********************************************

Hi there,
I have a question, if you have let's say 56 cores on a single workstation, and 2 users that would like to submit job to this workstation mainly via Windows, what would be the best practice for this?
should I set the workstation up as a linux server and install all ANSYS packages on it and have the 2 users submit their job from the windows using RSM or there are better ways to do this?
Thanks