Message boards :
Number crunching :
Near Time Persistency Checker (NTPCKR)
Message board moderation
Previous · 1 · 2 · 3 · Next
Author | Message |
---|---|
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
Henri et al Back in November 2007 I had a conversation with Eric. I was aware that Jeff/Matt were working on the coding, Eric stated they had managed to salvage hardware to cobble a machine for the proof of concept. The then hope was the Science Database would be ready and they would "fire it up" sometime this month. Adding the index's and other issues have slowed that a bit. So as Bob found a way to add the indexes without killing WorkUnit production it was step forward again. The other portion is (as Eric mentioned briefly) that with the data that needs to be gone through if the only machine that is running NITPCKR is the "cobbled togehter machine" it will take 3-4 years to go through the existing data (starting with the most recent MultiBeam). Thus Donations are very important (See Hardware Donations). As was already mentioned as Court left mid year; the three Matt, Eric and Jeff are the only ones minding the store. So, NTPCKR could be ready before the end of 2008? Cool! Thanks for the info, Pappa! :) Please consider a Donation to the Seti Project. |
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
I spoke with Eric today, I asked a pointed question about NTPCKR. I hoped that he would have had time to post about the progress... I recommended that Server status be shown in the Server Status page. This also would take some coding to show what progress is happening. At this point I have added to Matt's ToDo List... My Bad... Any PHP Coders around? Regards Al Please consider a Donation to the Seti Project. |
Logan Send message Joined: 26 Jan 07 Posts: 743 Credit: 918,353 RAC: 0 |
I spoke with Eric today, I asked a pointed question about NTPCKR. Good news, Pappa! Is only we want. To have acknowledge of the progress. Best Regards from Barcelona (Spain). Logan. BOINC FAQ Service (Ahora, también disponible en Español/Now available in Spanish) |
Josef W. Segur Send message Joined: 30 Oct 99 Posts: 4504 Credit: 1,414,761 RAC: 0 |
A Supermicro 6025W-NTR with 32 GB RAM in 4GB sticks, 2 Intel E5440 Harpertown 45 nm quads, and a few Terabytes of disk space can be had for under $7000 from various configurators. Here's a saved example from ServersDirect. The RAM could later be expanded to 64 GB using the other 8 slots, additional disk capacity is also easy to add. Is that level of computational power really needed? I don't know, but I can take on faith that the hardware donations request is the project's best guess. Joe |
Jason A. Countryman Send message Joined: 29 Aug 03 Posts: 139 Credit: 50,172,873 RAC: 2 |
For just over $4000 you can get a dell poweredge 1950 III server with dual E4520 quads, 32GB ram and 1.5 tb hd here |
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
Joe et al Yes that is the kind of power needed. This goes back to what I was told was cobbled together and the expected runtime for all the data in the Science database would be 4+ years. Because as NTPCKR gets running then any new result that is return is placed in the Top of the Que to see what is happening. As it was explained to me the model would be, the most recent data (which all the indexes in place) and New returning Data. Thus what we are doing now is very close to the top of the list. What has been processed starts working backwards to the beginning of time (the 1st Result of Seti classic). Unofficial: In a current conversation with Eric, Jeff has NTPCKR running. There are things that are waiting for an offical announcment and the machines addition to the server status page not to mention the coding progress pages and most likely candidates. Or the question, how do you build an elephant.... The only answer is from the ground up! A Supermicro 6025W-NTR with 32 GB RAM in 4GB sticks, 2 Intel E5440 Harpertown 45 nm quads, and a few Terabytes of disk space can be had for under $7000 from various configurators. Here's a saved example from ServersDirect. The RAM could later be expanded to 64 GB using the other 8 slots, additional disk capacity is also easy to add. Please consider a Donation to the Seti Project. |
Josef W. Segur Send message Joined: 30 Oct 99 Posts: 4504 Credit: 1,414,761 RAC: 0 |
For just over $4000 you can get a dell poweredge 1950 III server with dual E4520 quads, 32GB ram and 1.5 tb hd here I guess you meant E5420, and that's certainly a possibility. The master science database has about 1.6 billion signals and we added more than 2 million in the last 24 hours, suggesting the database will double in size within 2 years. I worry a bit that an adequate NTPCKR for now might be overwhelmed before the hardware is obsolete. That's why I was looking at systems with room for future upgrades. Joe |
Blurf Send message Joined: 2 Sep 06 Posts: 8962 Credit: 12,678,685 RAC: 0 |
UPDATE: From Matt's Daily Update: http://setiathome.berkeley.edu/forum_thread.php?id=46478 Astropulse and the ntpckr are both just around the corner as well. I know we've been saying that a while, but it's getting truer ever day. Lots of big things coming down the pike. |
Fred J. Verster Send message Joined: 21 Apr 04 Posts: 3252 Credit: 31,903,643 RAC: 0 |
UPDATE: Glad hearing this, Blurf and 'seeing' you're well, after this awfull incident. |
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
Matt stated in the Tech News that here over the next couple of days you might start seeing some of the "first" results from NTPCKR Please consider a Donation to the Seti Project. |
Matt Lebofsky Send message Joined: 1 Mar 99 Posts: 1444 Credit: 957,058 RAC: 0 |
Matt stated in the Tech News that here over the next couple of days you might start seeing some of the "first" results from NTPCKR Well.. not results as much as interesting data byproducts (see science status page). But definitely more and more as the nitpicker gets developed. - Matt -- BOINC/SETI@home network/web/science/development person -- "Any idiot can have a good idea. What is hard is to do it." - Jeanne-Claude |
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
Matt Thank You I see you have coined a new word "Multiplets LOL Matt stated in the Tech News that here over the next couple of days you might start seeing some of the "first" results from NTPCKR Al Please consider a Donation to the Seti Project. |
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
While I have been busy elsewhere work on NTPCKR continues... Matt's Post Hiccup Stars (May 28 2008) So finding little "nuggets" in the Tech News is where a part of this comes from. Please consider a Donation to the Seti Project. |
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
Progress still happens Matt's post to the Tech News More Bits (Jun 17 2008) A lot of what's been going on the past couple of days has been post-vacation catchup (half the staff was out of town). While I have a zillion other things to do I discovered a couple ways to optimize the NTPCkr so I coded that up and I'm testing it now. Every little speedup on this front helps. Jeff's still working on the scoring part. We're getting there... Please consider a Donation to the Seti Project. |
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
Other than a request for a Server that is capable of doing the job not much has been said. With Other problems on the server side this has no news... I Ill bring up back as I can get information Thank You Pappa Please consider a Donation to the Seti Project. |
Jon Golding Send message Joined: 20 Apr 00 Posts: 105 Credit: 841,861 RAC: 0 |
Any chance of a brief update on the NTPCkr? Last I heard decisions were being made on the optimum scoring algorithm - has that decision been made? Any hold-ups or show-stoppers? Is there any help that we volunteers can do in terms of Beta testing? Just curious. |
Jeff Cobb Send message Joined: 1 Mar 99 Posts: 122 Credit: 40,367 RAC: 0 |
After spending time on radar blanking and various maintenance chores, I'm getting back to the ntpckr. Here is a quick status report. The bulk of the code is written and fairly well tested. Next up - the hardest code. Still to do is the scoring code and some low level code that interfaces with Informix. It turns out that one of the Informix data types we like to use (SQLLIST) will not hold candidate lists of the size we want to store! We also need to integrate the Gnu Scientific Library with our software because the numerical code we historically have used (Numerical Recipes) does not allow inclusion in open source projects. There are a number of smaller tasks to do as well, such as getting star catalogs into our database mapped with a useful schema. I will try to get into the habit of posting here as we (myself, Eric, and Matt) complete various ntpckr development tasks. To answer Jon's specific questions, no final decisions yet on optimal scoring algorithms but work is getting done on that front. No show stoppers thus far! Thanks for asking about beta testing. Unfortunately, the ntpckr is dependent on fast read/write access to the science database which is not feasible beyond the lab machines. Jeff |
Gary Charpentier Send message Joined: 25 Dec 00 Posts: 30639 Credit: 53,134,872 RAC: 32 |
Thanks for the info Jeff. Look forward to hearing more about the progress. Gary |
speedimic Send message Joined: 28 Sep 02 Posts: 362 Credit: 16,590,653 RAC: 0 |
Thanks for the update, Jeff! I will try to get into the habit of posting here as we (myself, Eric, and Matt) complete various ntpckr development tasks. Just thinking if the Technical News would be a better place to post. Some posts tend to get lost among the masses here... mic. |
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
For ages I was the only one that looked (for) to dig out/report information about NTPCKR. Here is where people decided they wanted to read it, not the Tech News "we never go there." It was one of those things about "why I will not donate as there is no progress." With other problems taking the forefront (currently) and nothing new for a while I let it go. I knew there would be a time when we had News. Now with the Hardware Donations Thread, part of what is stated is that testing on the Cobbled Together Server (or the server parts that Matt and Eric could get to run) they are at the point they need the Real Machine. Eric told me ages ago that only using the listed machine it would only take about three years to get through the Master Science Database. Or the simplistics of freshly returning workunits would get handled more immediately and the older ones would take longer. Like Flying over the Atlantic Ocean and dropping a "single result" from 40,000 feet then looking for the ripples. Thanks for the update, Jeff! Please consider a Donation to the Seti Project. |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.