CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > General Forums > Hardware

$2000 desktop: suggestions?

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   September 5, 2014, 00:09
Question $2000 desktop: suggestions?
  #1
New Member
 
Freewill1's Avatar
 
Join Date: Aug 2014
Posts: 18
Rep Power: 11
Freewill1 is on a distinguished road
Hi all,
I'm new to CFD-Online and have been lurking around here and learning. I'm a student working on geodynamics and dealing with numerical modeling problems like mantle convection, plates collision/subduction, etc.

Here is an typical 2-D case I from paper[1], which is to evaluate the burial and exhumation processes of deep rocks in Alpine-type orogens. To do this, a specialised FEM code named SOPALE-Nested is used. The features is listed as following:

(1) the thermal–mechanical creeping (stokes) flows at the crust or mantle scale is solved;

(2) model materials deform by frictional-plastic (brittle) or power-law viscous (ductile) flow that includes respective strainsoftening/strain-weakening mechanisms;

(3) as for the scale of the model, the domain is 2000×660 km with resolution of 1×1 km.

Here is several figures and animations of the model setups and results:





animation 1: here

animation 2: here


A important thing I learned is that a bandwidth between CPU and RAM really matters for CFD application, and as many people suggested, the newst chip i7-4790K($294.29 on Amazon) has only 2 memory channels but the i7-4820k($319.99 on Amazon) has four and performs better. For me, due to limited budget, I can only afford a system less than $2000.

So I have a question of what kind of system do you think I should buy to meet the need? Many thanks for suggestions of what to buy: CPU/RAM/motherboard/graphic card/...
or do you think a HP Z420 with intel xeon E3 CPU will be good choice?

Best,
Freewill

References:
[1] Butler, J.P., Beaumont, C. and Jamieson, R.A., 2013. The Alps 1: A working geodynamic model for burial and exhumation of (ultra)high-pressure rocks in Alpine-type orogens. Earth and Planetary Science Letters, 377–378(0): 114-131.
Freewill1 is offline   Reply With Quote

Old   September 5, 2014, 05:21
Default
  #2
Super Moderator
 
flotus1's Avatar
 
Alex
Join Date: Jun 2012
Location: Germany
Posts: 3,395
Rep Power: 46
flotus1 has a spectacular aura aboutflotus1 has a spectacular aura about
With a budget of $2000 you might even consider the new Haswell-E processors like I7-5820K or 5930K. They also have 4 memory channels and support DDR4 ram.

Edit:
An even better solution in terms of memory bandwidth and cost efficiency might be a setup with 2 AMD Opteron processors.
This will almost have twice the memory bandwith compared to the Intel system and a much higher number of cores.
The performance per core will be lower than for the Intel system, but If you are not limited in the number of cores you can use for your simulation it is worth considering.

Last edited by flotus1; September 5, 2014 at 07:57.
flotus1 is offline   Reply With Quote

Reply

Tags
creeping flow, geodynamics, stokes flow, xeon

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Desktop Selection richardchen Hardware 10 September 3, 2014 09:36
Hardware build review/advise for student desktop AerE Hardware 10 July 5, 2013 07:03
[ICEM] Bocking Strategy Suggestions for 2d Trefoil Shape Inside Rectangle mullzer ANSYS Meshing & Geometry 2 March 2, 2012 15:06
Unforeseen problems in scaling up a cluster built with desktop parts? kyle Hardware 22 January 18, 2012 14:46
suggestion need desktop configuration for fluent Ravinder Reddy FLUENT 3 September 9, 2005 09:09


All times are GMT -4. The time now is 01:54.