Mersenne Digest Sunday, May 6 2001 Volume 01 : Number 846 ---------------------------------------------------------------------- Date: Sat, 31 Mar 2001 07:28:05 -0600 From: "Richard B. Woods" <[EMAIL PROTECTED]> Subject: Re: Mersenne: Countdown is slowing In reply to fay aron charles <[EMAIL PROTECTED]>: > It shows that the countdown to testing all exponents below M(6972593) > at least once is 33. I see this a very large milestone. Last week the > countdown was 34, and the week before that it was 34. Patience. All earlier milestones were passed, and this one will be achieved and passed also. > I would assume the slow progress is because the last few exponents are > reserved by people with very slow processors who still update with the > server at least every 60 days. I have one of those exponents reserved, and am actively conducting its L-L test on my P75 even as I read your message and compose this response. With this processor I have completed L-L testing on more than 30 Mersenne numbers (in addition to factoring work on about that many others) since I joined GIMPS. Never have I delayed the achievement of a GIMPS milestone. Never -- not once -- have I quit any test without immediately properly checking the exponent back in. In a few cases, I stopped testing an exponent and checked it back in because a revision in my private theory about the occurrence of Mersenne primes indicated that there was no chance for that candidate to be prime. Why do I participate? The slim chance of fame in case I discover a Mersenne prime. Even though my private theory about the occurrence of Mersenne primes predicts that exponents other than the ones I have reserved through GIMPS have higher chances of being prime, I continue to cooperate with GIMPS by working with exponents I am assigned, instead of "poaching" by testing exponents that are assigned to other participants. > Could these last few exponents be assigned for double checking now? > It would at least satisfy my curiosity. ... and, by allowing someone with a much faster processor to L-L test my assigned exponent, it would also allow that someone to determine whether the number under test is prime, or not, _before I do_. I.e., your proposal would rob me of my chance for fame by breaking the current contract that my reservation means that only I have the opportunity for potential fame in discovering that "my" M(6xxxxxx) is prime, as long as I fulfill my part by making reasonable progress. If that were allowed, then why should I continue contributing to GIMPS at all ... ... when my just-arrived Athlon 1.2 GHz (that's 16 * 75 MHz) processor is ready to join in? If GIMPS breaks its contract with me, I'll take my new "fast" processor elsewhere. _________________________________________________________________________ Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers ------------------------------ Date: Wed, 02 May 2001 07:42:15 -0700 From: "Gary Untermeyer" <[EMAIL PROTECTED]> Subject: Re: Mersenne: Countdown is slowing Greetings, Hey Richard! I applaud your patience and perserverence. It's not so much how fast a GIMPS computer is, but how well its owner keeps it "on track." Milestones would be reached much quicker if all had the same attitude toward responsibility. I currently have over 300 exponents tested, many of them done on a P100. It's still doing double checking work while my new P733 (well, relatively new) cranks away on other things for GIMPS. Keep on truckin', Gary _________________________________________________________________________ Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers ------------------------------ Date: Sun, 6 May 2001 18:26:21 -0600 From: "Matt Goodrich" <[EMAIL PROTECTED]> Subject: Mersenne: Slow CPU's in a Proliant 2500 I have just upgraded a Proliant 2500 from dual PPro 200's to dual Pentium II 333 overdrive processors. According to Intel's website this is supported. Also Compaq offers this upgrade. Now before I upgraded, I was running double check's on 2 exponents, 668xxxx. I was getting about .515 second iteration times. Now that I have upgraded, I am only getting .448 second iteration times. Yes, the affinity is set properly. I have set the switches on the I/O board for 66/333 speed. I believe my problem stems from having the wrong ROM chips. The server, when booting, says the BIOS is version E24 which is associated with the PPro 200 processors. I went to Compaq's web site and found that the E50 version is for PII 333's. Downloaded the ROMPaq but it will not flash. It says there isn't a valid ROM Image on the floppy disk for the installed device. After searching Compaq's web site (with their lousy search engine I might add) I found an article that say's I probably need to replace the physical ROM chips. Now for my question. Has anyone here done this with this server (upgraded processors and had to change the ROM chips)?? I want to know if this will fix the problem, or is there something else going on? Interesting note here. If I run just 1 copy of Prime on either CPU I get .325 second iteration times. It doesn't matter which copy of the program or which CPU I run it on. It is only when I run both copies that it slows down to .448 seconds. BTW I am running Windows 2000 Server with the latest SSD from Compaq. This is what I have done to the server to try and resolve this: 1.) Ran Smart Start and used the system erase utility, then reconfigured the server, drive array etc. 2.) Reloaded Windows 2000 Server. 3.) Reinstalled Prime (both instances) and then brought my saved work back in. 4.) Flashed the BIOS to the newest version for the E24. I was hoping this might change something and let me flash with the ROMPaq I need, but it didn't work out that way. 5.) I also have stopped all unneeded services running in the background, such as IIS, Compaq agents, etc. Do you think Windows 2000 could just be stealing this many cpu cycles for itself? This box is not in a production environment. It is in my home. It isn't running Active Directory either. When I run just 1 copy though, task manager does show 3 to 7% activity on the idle processor. Hmmmmmm, maybe I should try NT Server 4.0 and see what happens. Any thoughts or suggestions would be appreciated. Matt _________________________________________________________________________ Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers ------------------------------ End of Mersenne Digest V1 #846 ******************************
