eOn to go BOINC! in 08-09
AMDave opened the dialog:
Quote:
Since 6-Sept the project turnover has almost flat-lined in spite of the available member resources.
(That seems to coincide with some nasty weather events in the state. Hope you are all ok)
We did read this:
Quote:
# Sept. 16, 2008Our recent paper is published. Thanks for everyone for their help with this research.
# Aug 5,2008: We have been using EON to test some improvements to our algorithm. This is working very well now, and the results are being published in a new paper to appear in the Journal of Chemical Physics. Here is a preprint. Thanks for everyone for their help with this research.
Congratulations on the paper.
I noted reference 43 on page 9.
I would like to know if there is more chem-work coming, if the project is still funded, or is it now in end-stage?
Admins - what's up?
Graeme responded:
Quote:
Fortunately, we are pretty safe from storms in Austin being so far inland. Ike didn't even bring rain here.
Our server, on the other hand, is less fortunate. It was in good hands with my postdoc, Lijun, and I think this new paper is a big step forward in the simulation methodology.
A new student, Rye, is taking over. We've purchased a new server to help cut down on the memory and communication problems. But I think we're settled on switching to boinc to solve our ongoing server issues. In that system, the issuing of work units is separated from any other calculations on the server, so that it is always responsive to client requests.
Our first new project will be to look at dynamics in high-temperature Ni alloys. These will be more expensive calculations, so we'll have to rescale the statistics.
In the long-term, we'll be switching the entire system to quantum calculations. There are still some fundamental issues to work out, but this will allow for modeling reaction dynamics at surfaces. This has great potential for designing new catalysts, particularly for energy conversion applications such as in fuel cell.
AMDave jumped for joy and said:
Quote:
Thank you very much for your kind reply, Graeme.
It is good to know that the project team fared well and that the project is looking to the future.
Please pass on our thanks to Lijun for the hard work and our welcome to Rye.
Rye may face a challenge in the raising of the eOn@Home BOINC platform and the integration of eOn into the framework, but should take heart in the knowledge that there are now many with BOINC server experience who are able to and willing to help.
Best of luck, Rye!
Don't be afraid to lean on us.
I do like the sound of the "more expensive" work units. That should indeed cut down the I/O per computer and will help suit the almost-batch-like BOINC IO interface.
It will be interesting to see how the project handles the insurgence of thousands of avid BOINC addicts (it is like being slash-dotted!). There is no shortage of keen testers with weird and wonderful platforms who will find and challenge the eOn@Home url as soon as it lands on the DNS servers.
I am very pleased that you mention the long-term refinement in the focus of the science. It follows on from some posts you made in that regard a couple of years ago, so it reasserts your long-term view in your line of study.
So it seems that you have some hard work and challenging days ahead.
That is excellent!
http://www.free-dc.org/forum/images/smilies/clap.gif
My best to the eOn project team for 08-09.
AMDave
www.amdusers.com
eOn to go BOINC! in 08-09
Quote:
Originally Posted by
vaughan
Thank you AMDave. That is very good news. Doesn't matter if its Boinc or not but I have always liked the EON project. The dialogue you have cut and pasted made for interesting and positive reading.
I agree - good luck to Graeme and his new team.
The CPUs are waiting. :)
Me Too :icon_mrgreen:
