FastFieldSolvers Forum
FastFieldSolvers Forum
Home | Profile | Register | Active Topics | Members | Search | FAQ
 All Forums
 FastFieldSolvers
 FastHenry2
 Great problem

Note: You must be registered in order to post a reply.
To register, click here. Registration is FREE!

Screensize:
UserName:
Password:
Antispam question: What do MOONwalk and MOONdance have in common?
Answer:
Format Mode:
Format: BoldItalicizedUnderlineStrikethrough Align LeftCenteredAlign Right Horizontal Rule Insert HyperlinkInsert EmailInsert Image Insert CodeInsert QuoteInsert List
   
Message:

* HTML is OFF
* Forum Code is ON

 
   

T O P I C    R E V I E W
Alejandro Posted - Mar 08 2014 : 14:53:54
Hello,

i'm using FastHenry through a Matlab script to accelerate the process. It works fine but i've realised that when you design the coil for a high number of turns, unless nwinc=1 and nhinc=1, message Out of memory appears. It's very serious because it seems impossible to obtain with FastHenry the resistance of a coil with a high number of turns.

Kind regards.
10   L A T E S T    R E P L I E S    (Newest First)
Enrico Posted - Sep 23 2015 : 17:58:19
You might be interested in knowing that we offer now the 64 bits version of both FastHenry2 and FastCap2.

On the 'Download' page of our web site you can find the new version 4.6.0 of the FastFieldSolver bundle. The 64 bits bundle now contains also the 64 bits version of FastHenry2 and of FastCap2, allowing to overcome the 4Gb memory limit of the previous 32 bits versions. Please download the package and update your installation.

For completeness, I also add that under the 'Download' section you can find the full sources for FastHenry2 ported under MS Visual Studio 2013 Community Edition, that is free for Open Source project (check detailed license conditions on the MS site), supporting both 32 bits and 64 bits compilation and with zero warnings. The sources are now hosted on GitHub, again see details under the 'Download' section.

Best Regards,
Enrico
Alejandro Posted - May 13 2014 : 17:07:57
Thank you but for now our budget for the project is zero :(. But thanks for the offer, i'll take it into account.
Enrico Posted - Apr 27 2014 : 06:52:20
I'm sorry, but there is no x64 version of FastHenry in the short term. We plan an integration into FasterCap in the medium term, so you get only one tool for impedance extraction, but it is not ready now.

We could prepare a x64 Win porting of FastHenry, but it will have some unforeseen effort associated, so we could do it as a service, if you are interested; but I'm afraid that in this case it can't be for free, sorry about that. Just tell me if you are interested.

Best Regards,
Enrico
Alejandro Posted - Apr 15 2014 : 13:39:26
Sorry,

i'm back :(. People from Whiteley answered me that no x64 version exists for windows.

Could you recommend me a solution? Using linux perhaps, or migrating linux version to windows?

Thank you again.
Alejandro Posted - Apr 06 2014 : 18:18:23
Thank you, i'll try it.

Best regards.
Enrico Posted - Mar 19 2014 : 23:16:18
As a quick solution, Whiteley Research might have a x64 version for Windows, although without GUI and FastModel integration. You can check on wrcad.com

Best Regards,
Enrico
Alejandro Posted - Mar 16 2014 : 18:26:46
Initially it is not an option, because unfortunately i haven't used before Matlab, automation and excel, which are the tools involved, in Unix :(.

If a version for 64b existed, it would be great.

Thanks.
Enrico Posted - Mar 13 2014 : 09:52:20
Ok so my offline hypothesis was correct, you hit the memory limit. This is mainly imposed by the Operating System (in principle, with 32 bits architectures you could be able to use 4G Bytes).

As a shortcut solution you might use the *nix version of FastHenry. Please let me know if this is an option for you.

Best Regards,
Enrico
Alejandro Posted - Mar 11 2014 : 19:55:18
Thank you very much.

For example, simulating 2 square coils of 100 turns, with segments of rh=rw=2 and nwinc=nhinc=5 I get:

fasthenry: out of memory in file `D:\Documenti\Codice\sample\FastHenry332_110325\FastHenry\direct.c' at line 65
(NULL pointer on 20000 byte request)
Total Memory Allocated: 1446246 kilobytes (brk = 0x4575a4)
Q2M matrix memory allocated: kilobytes
Q2L matrix memory allocated: kilobytes
Q2P matrix memory allocated: 1051953 kilobytes
L2L matrix memory allocated: kilobytes
M2M matrix memory allocated: kilobytes
M2L matrix memory allocated: kilobytes
M2P matrix memory allocated: kilobytes
L2P matrix memory allocated: kilobytes
Q2PD matrix memory allocated: 341845 kilobytes
Miscellaneous mem. allocated: 2834 kilobytes
Inductance mem. allocated: 49613 kilobytes
Total memory (check w/above): 1399467 kilobytes


I have 16GB of RAM and more than 100 free GB in hard disk. I'd like to ask you if i can solve it anyhow.

Regards.

Enrico Posted - Mar 10 2014 : 22:19:27
From the information you provide, it is difficult to help you. Of course there is a memory limit in the dimension of the structures you can simulate, and this is however depending on the amount of memory you have and on the fact that FastHenry, on Windows, is a 32 bit process, so you are limited to about 2GBytes of memory.
If you enable the "Print debugging information (-D)" option you will have a summary of the allocated memory, and can realize, increasing the number of turns and/or the complexity of the structure, how close you are to the memory limit.

Best Regards,
Enrico


FastFieldSolvers Forum © 2020 FastFieldSolvers S.R.L. Go To Top Of Page
Powered By: Snitz Forums 2000 Version 3.4.06