Double checking - Page 5

gd_barnes · 2007-11-26, 17:42

Kosmaj,

The double-check effort that you requested for k=213, 237, 249, 261, 267, 273, 279, & 291 has finished sieving and started LLRing 2-3 days ago. It has completed n=100K-160K so far. No problems yet.

The next status will be when I'm done (3-4 weeks) or earlier if I find a problem.

Gary

kar_bon · 2007-11-29, 14:00

the Prime Search pages has been extended for all k<255 downto k=3!
perhaps another doublecheck effort when they in higher ranges, mostly for the new k's at n=140.000.
karsten

Kosmaj · 2007-12-02, 07:07

I tested all gaps of k=285 marked on the Keller's page:
135-145, 155-170 (I did 135-173 including 145-155), and 230-260. I confirmed 3 primes in the 145-160 range but to my amazement I found two new primes

285 161183
285 163071

They were not supposed to be there because the range was tested by somebody from the 15k project (I don't know who because Joss didn't leave any records) and by a guy called Footmaster from the PrimeSearch project but they either used one of early, faulty versions of LLR or they just didn't report them. Anyway, k=285 is now complete to 808k without any gaps.

I'm now working on the 100-200k gap of k=181, and currently I'm at 150k, no surprises here so far, I confirmed two primes at 132 and 133k.

All primes mentioned above are listed on our k<300 stats page.

em99010pepe · 2007-12-08, 10:24

I have a question. I notice my quad-core with the stock fan was getting too damm hot so I decided the clean the dust. Temperatures went down by 20ºC. My problem, maybe not a problem, is that my last three primes were found with the default core between 70-80 ºC. Do I need to double-check all my ranges? I know from Ars forum that temps can go high as 90 ºC with super coolers and with the cpu overclocked. Comments please.

Carlos

em99010pepe · 2007-12-08, 11:32

Having bad results would be only with half of the 105 ranges, two 5th ranges, and two ranges from 125, where the primes were found! I think it's not necessary to double-check but of course I'll wait for the answer of the experts.

mdettweiler · 2007-12-08, 17:10

Quote:

Originally Posted by em99010pepe

Having bad results would be only with half of the 105 ranges, two 5th ranges, and two ranges from 125, where the primes were found! I think it's not necessary to double-check but of course I'll wait for the answer of the experts.

Although I am by no means an expert on this, I would guess that the ranges for which the primes were found in are probably good--since all primes submitted to the Prime Pages are automatically re-checked.

A while ago I'd had some problems with LLR crashing, and when I cleaned out my case, it fixed the problem of LLR crashing, as well as giving me a nice speed boost (more than 2x). From what I can tell, though, I wasn't producing bad residuals anywhere in there--during part of the time while I had the LLR-crashing problem, I was still running some PrimeGrid LLR-TPS workunits through BOINC; 50% of their LLR-TPS workunits are set with a quorum of 2 (i.e. they are automatically doublechecked), and I wasn't getting any validate errors, so thus I'm guessing that I most likely didn't produce any bad residuals. As for your quadcore, considering that everything is still running stably (I'm assuming that it is), most likely your residuals were fine, just that maybe your system wasn't running at top speed (as I mentioned before, mine got a big speed boost when I cleaned it out).

Just my $0.02.

em99010pepe · 2007-12-08, 17:32

Didn't get a speed boost...the machine didn't crash...everything very stable...I am impressed.

AES · 2007-12-08, 19:48

Carlos,

I found this data from Intel (on tomshardware.com) regarding acceptable temp ranges based on the core stepping the quad uses.

http://mersenneforum.org/showpost.ph...7&postcount=10

em99010pepe · 2007-12-08, 21:01

Thanks, now it's running super cool. Core 0 at 55ºC and the rest at 49ºC at full load. Also I'm waiting for a new cooler.

VBCurtis · 2007-12-09, 18:04

Anon is right- all primes submitted to top-5000 are double-checked, so you have the reassurance that all your primes were confirmed as a substitute for double-checking.

The Core2 architecture has a LOT of headroom built-in. 70C temps should present no problems at stock speeds; of course, cleaning out the dust often makes everything more reliable.

The stock cooler is plenty for non-overclocked machines. A larger cooler will also collect dust, and need cleaning only slightly less often than your old one.
-Curtis

gd_barnes · 2007-12-20, 01:36

Kosmaj,

The double-check effort that you requested for k=213, 237, 249, 261, 267, 273, 279, & 291 has completed. No problems were found.

Ranges will be reported to Wilfred Keller.

Gary

2007-12-02, 07:07	#47
Kosmaj Nov 2003 E26₁₆ Posts	k=285 I tested all gaps of k=285 marked on the Keller's page: 135-145, 155-170 (I did 135-173 including 145-155), and 230-260. I confirmed 3 primes in the 145-160 range but to my amazement I found two new primes 285 161183 285 163071 They were not supposed to be there because the range was tested by somebody from the 15k project (I don't know who because Joss didn't leave any records) and by a guy called Footmaster from the PrimeSearch project but they either used one of early, faulty versions of LLR or they just didn't report them. Anyway, k=285 is now complete to 808k without any gaps. I'm now working on the 100-200k gap of k=181, and currently I'm at 150k, no surprises here so far, I confirmed two primes at 132 and 133k. All primes mentioned above are listed on our k<300 stats page.

2007-12-08, 10:24	#48
em99010pepe Sep 2004 2×5×283 Posts	I have a question. I notice my quad-core with the stock fan was getting too damm hot so I decided the clean the dust. Temperatures went down by 20ºC. My problem, maybe not a problem, is that my last three primes were found with the default core between 70-80 ºC. Do I need to double-check all my ranges? I know from Ars forum that temps can go high as 90 ºC with super coolers and with the cpu overclocked. Comments please. Carlos Attached Thumbnails Last fiddled with by em99010pepe on 2007-12-08 at 10:25

2007-12-20, 01:36	#55
gd_barnes "Gary" May 2007 Overland Park, KS 11819₁₀ Posts	double-check for 8 k's 213<=k<=291 complete Kosmaj, The double-check effort that you requested for k=213, 237, 249, 261, 267, 273, 279, & 291 has completed. No problems were found. Ranges will be reported to Wilfred Keller. Gary

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Double checking of Results	pinhodecarlos	Prime Gap Searches	13	2017-12-09 06:07
What about double-checking TF/P-1?	137ben	PrimeNet	6	2012-03-13 04:01
Double checking	Unregistered	Information & Answers	19	2011-07-29 09:57
Double-checking milestone?	jobhoti	Math	17	2004-05-21 05:02
Any glory in double checking?	Quacky	Lounge	5	2003-12-03 02:20

2007-11-26, 17:42	#45
gd_barnes "Gary" May 2007 Overland Park, KS 53·223 Posts	Kosmaj, The double-check effort that you requested for k=213, 237, 249, 261, 267, 273, 279, & 291 has finished sieving and started LLRing 2-3 days ago. It has completed n=100K-160K so far. No problems yet. The next status will be when I'm done (3-4 weeks) or earlier if I find a problem. Gary

2007-11-29, 14:00	#46
kar_bon Mar 2006 Germany 2999₁₀ Posts	the Prime Search pages has been extended for all k<255 downto k=3! perhaps another doublecheck effort when they in higher ranges, mostly for the new k's at n=140.000. karsten

2007-12-08, 11:32	#49
em99010pepe Sep 2004 2830₁₀ Posts	Having bad results would be only with half of the 105 ranges, two 5th ranges, and two ranges from 125, where the primes were found! I think it's not necessary to double-check but of course I'll wait for the answer of the experts.

2007-12-08, 17:32	#51
em99010pepe Sep 2004 2×5×283 Posts	Didn't get a speed boost...the machine didn't crash...everything very stable...I am impressed.

2007-12-08, 19:48	#52
AES Jul 2007 Tennessee 260₁₆ Posts	Carlos, I found this data from Intel (on tomshardware.com) regarding acceptable temp ranges based on the core stepping the quad uses. http://mersenneforum.org/showpost.ph...7&postcount=10

2007-12-08, 21:01	#53
em99010pepe Sep 2004 2·5·283 Posts	Thanks, now it's running super cool. Core 0 at 55ºC and the rest at 49ºC at full load. Also I'm waiting for a new cooler.

2007-12-09, 18:04	#54
VBCurtis "Curtis" Feb 2005 Riverside, CA 2²·1,409 Posts	Anon is right- all primes submitted to top-5000 are double-checked, so you have the reassurance that all your primes were confirmed as a substitute for double-checking. The Core2 architecture has a LOT of headroom built-in. 70C temps should present no problems at stock speeds; of course, cleaning out the dust often makes everything more reliable. The stock cooler is plenty for non-overclocked machines. A larger cooler will also collect dust, and need cleaning only slightly less often than your old one. -Curtis