|
[Sponsors] |
October 7, 2019, 08:03 |
Workstation for particle-based methods
|
#1 |
New Member
Miguel Molinos
Join Date: Dec 2015
Location: Madrid
Posts: 14
Rep Power: 11 |
Dear all,
The research group where I am is planning to acquire a new workstation for simulate 3D particle-based simulations, SPH or MPM for instance. We are planning also to employ HPC techniques for the calculations such as CUDA to accelerate the computations, so we will need CUDA-capable NVIDIA graphics cards. Our budget is about 10.000 €. Do you think it is reasonable? Do you have any suggestions about GPU, CPU, motherboard or cooling system? Thanks in advance Miguel Last edited by migmolper; October 7, 2019 at 10:23. |
|
October 7, 2019, 16:55 |
|
#2 |
Super Moderator
Alex
Join Date: Jun 2012
Location: Germany
Posts: 3,428
Rep Power: 49 |
So what exactly are you aiming for?
Do your codes scale across multiple GPUs? Do you want NVLink? Do you need single or double precision compute performance? How much VRAM do you need? Are you 100% sure that GPU acceleration is the way to go over CPU? I know, these are a lot of questions. But For GPU acceleration to be worth it with a limited budget for hardware, you need to be sure about these details. And if GPU acceleration is viable, it is very important to choose the right one for the application. Last edited by flotus1; October 8, 2019 at 03:11. |
|
Tags |
cuda, hpc, particle methods, workstation |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Airfoil Shape Optimization using gradient based methods | jaiswal | Main CFD Forum | 1 | April 10, 2014 12:22 |
comments on FDM, FEM, FVM, SM, SEM, DSEM, BEM | kenn | Main CFD Forum | 2 | July 18, 2004 19:28 |
Workstation recommendations | Charles Crosby | Main CFD Forum | 5 | June 21, 2000 05:43 |
PC vs. Workstation | Tim Franke | Main CFD Forum | 5 | September 29, 1999 16:01 |
pressure correction methods | clifford bradford | Main CFD Forum | 4 | August 23, 1999 12:42 |