Mathematica Kernel Stops Functioning

[john...@gmail.com]

I installed Mathematica 9 on my computer (64 bit quad core AM 3450 (AMD) processor with 6 GB of ram) about a week ago. I've run a couple of notebooks (a notebook with intensive numerical computation and a notebook with intensive symbolic computation) which I developed on Mathematica versions 6, 7, and 8. Both of these notebooks work very well with Mathematica 6, 7, and 8.

However, Mathematica 9 starts off with great fervor (as I watch with the Windows 7 (64 bit) Resource Monitor), but after about an hour of computation, the Mathematica kernel simply stops working and the CPU runs at only about 1% apparently performing only inscrutable PC housekeeping tasks. Not surprisingly, the notebook never ends, and I get no result.

Does anyone have an idea why this is happening? Or better yet, how to fix it?

Thanks,

John

[David Bailey]

You are going to have to supply more evidence, or narrow the issue down
somewhat.

Installing Mathematica 9 will not remove previous versions, so I would
try running your problems on a previous version just to confirm that
they do still work.

Obviously, if the notebook is reasonably self contained, it may be best
to just send it to Wolfram Support, but I'd start by watching to
determine which cell of your notebook actually hangs.

It may also help to add a few Print statements at likely places to
follow the progress of your calculation.

If the notebook is small and self-contained enough, you could put it
somewhere on the internet and post a link here.

David Bailey
http://www.dbaileyconsultancy.co.uk

[john...@gmail.com]

David,

Thanks for your reply. I've made a notebook with the algorithm and a data set, the analysis program I've written, and the results using

Mathematica 8 which plots 43 data points(3100 seconds to completion with Ubuntu Linux, 32 bit 2.4 GHz Intel quad core processor)

and using

Mathematica 9 plotting only 7 data points(8000 seconds to completion with with Windows 7, 64 bit 3.6 GHz AMD quad core processor).

I've sent the notebook to Mathematica support and they are trying to understand the gross performance difference.

In the mean time, if you would like to fiddle around with it, I could send you the notebook (~1.1 M bytes, virtually all of which is a data set).

John Feth

[Thomas P. Melehan]

Please share the results of your discussions with Wolfram - I have experienced much the same problem..

Thanks

Tom

On Jul 23, 2013, at 5:21 PM, john...@gmail.com wrote:

> On Sunday, July 21, 2013 1:13:55 AM UTC-7, David Bailey wrote: