Other

drummers-lowrise

Message boards : Number crunching : Sier"pi"nski's Birthday Challenge

 Subscribe SortOldest firstNewest firstHighest rated posts first
Author Message
Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149196 - Posted: 6 Mar 2021 | 20:26:26 UTC

Welcome to the Sier"pi"nski's Birthday Challenge

The second challenge of the 2021 Series will be a 10-day challenge celebrating the birthday of Wacław Franciszek Sierpiński, Polish mathematician and namesake of the Sierpinski Problem, known to PrimeGrid as Seventeen or Bust. The challenge will be offered on the SOB-LLR application, beginning 14 March 12:00 UTC and ending 24 March 11:59 UTC.

Most people celebrate March 14th as Pi Day. Now, there are a myriad of arguments for why pi is undeserving of all the attention of this holiday, not the least of which is the fact that all the pi hubbub overshadows the birthday of a real mathematical hero: Wacław Franciszek Sierpiński. (Come on, he even has "pi" in his name.)

Sierpinski was born on March 14th, 1882 in Warsaw. He has his own street in Warsaw, a statue at Cambridge University, and even a crater on the moon named after him. The extraordinary Polish mathematician was honored with all these due to his pioneering work on fractals, contributions to set theory, and his aid in cracking Soviet army codes. Most relevant to us here at PrimeGrid is his work in number theory where we find the Sierpinski problem, which is what the Seventeen or Bust project has devoted over 18 years and 6 million PetaFLOPS toward solving.

For more about Sierpinski and his Problem, check out the Seventeen or Bust synopsis below, or this fantastic article.

To participate in the Challenge, please select only the Seventeen or Bust LLR (SOB) project in your PrimeGrid preferences section.

Note on LLR2 tasks: LLR2 has eliminated the need for a full doublecheck task on each workunit, but has replaced it with a short verification task. Expect to receive a few tasks about 1% of normal length.

Application builds are available for Linux 32 and 64 bit, Windows 32 and 64 bit and MacIntel. Intel and recent AMD CPUs with FMA3 capabilities (Haswell or better for Intel, Zen-2 or better for AMD) will have a very large advantage, and Intel CPUs with dual AVX-512 (certain recent Intel Skylake-X and Xeon CPUs) will be the fastest.

Note that LLR is running the latest AVX-512 version of LLR which takes full advantage of the features of these newer CPUs. It's faster than the previous LLR app and draws more power and produces more heat, especially if they're highly overclocked. If you have certain recent Intel Skylake-X and Xeon CPUs, especially if it's overclocked or has overclocked memory, and haven't run the new AVX-512 LLR before, we strongly suggest running it before the challenge while you are monitoring the temperatures.

Multi-threading is supported and IS recommended. (SOB tasks on one CPU core will take 5-10 days on fast/newer computers and 2 weeks+ on slower/older computers.)

Those looking to maximize their computer's performance during this challenge, or when running LLR in general, may find this information useful.

• Your mileage may vary. Before the challenge starts, take some time and experiment and see what works best on your computer.
• If you have a CPU with hyperthreading or SMT, either turn off this feature in the BIOS, or set BOINC to use 50% of the processors.

• If you're using a GPU for other tasks, it may be beneficial to leave hyperthreading on in the BIOS and instead tell BOINC to use 50% of the CPU's. This will allow one of the hyperthreads to service the GPU.

• The new multi-threading system is now live. Click here to set the maximum number of threads. This will allow you to select multi-threading from the project preferences web page. No more app_config.xml. It works like this:

• In the preferences selection, there are selections for "max jobs" and "max cpus", similar to the settings in app_config.
• Unlike app_config, these two settings apply to ALL apps. You can't chose 1 thread for SGS and 4 for SoB. When you change apps, you need to change your multithreading settings if you want to run a different number of threads.
• There will be individual settings for each venue (location).
• The hyperthreading control isn't possible at this time.
• The "max cpus" control will only apply to LLR apps. The "max jobs" control applies to all apps.

• If you want to continue to use app_config.xml for LLR tasks, you need to change it if you want it to work. Please see this message for more information.
• Some people have observed that when using multithreaded LLR, hyperthreading is actually beneficial. We encourage you to experiment and see what works best for you.

Time zone converter:

The World Clock - Time Zone Converter

NOTE: The countdown clock on the front page uses the host computer time. Therefore, if your computer time is off, so will the countdown clock. For precise timing, use the UTC Time in the data section at the very top, above the countdown clock.

Scoring Information

Scores will be kept for individuals and teams. Only tasks issued AFTER 14th March 2021 12:00 UTC and received BEFORE 24th March 2021 11:59 UTC will be considered for credit. We will be using the same scoring method as we currently use for BOINC credits. A quorum of 2 is NOT needed to award Challenge score - i.e. no double checker. Therefore, each returned result will earn a Challenge score. Please note that if the result is eventually declared invalid, the score will be removed.

At the Conclusion of the Challenge

ABORTING tasks allows them to be recycled immediately; thus a much faster "clean up" to the end of an LLR Challenge. DETACHING, RESETTING, and PAUSING tasks causes them to remain in limbo until they EXPIRE. Therefore, we must wait until tasks expire to send them out to be completed.

Please consider either completing what's in the queue or ABORTING them. Thank you. :)

What is Seventeen or Bust?

Seventeen or Bust was a distributed computing project attempting to solve the Sierpinski problem. The name of the project is due to the fact that, when founded, there were seventeen values of k < 78,557 for which no primes were known.

Basically, the Sierpinski problem is "What is the smallest Sierpinski number?"

First we look at Proth numbers (named after the French mathematician François Proth). A Proth number is a number of the form k*2^n+1 where k is odd, n is a positive integer, and 2^n>k.

A Sierpinski number is an odd k such that the Proth number k*2^n+1 is not prime for all n. For example, 3 is not a Sierpinski number because n=2 produces a prime number (3*2^2+1=13). In 1962, John Selfridge proved that 78,557 is a Sierpinski number...meaning he showed that for all n, 78557*2^n+1 was not prime.

Most number theorists believe that 78,557 is the smallest Sierpinski number, but it hasn't yet been proven. In order to prove it, it has to be shown that every single k less than 78,557 is not a Sierpinski number, and to do that, some n must be found that makes k*2^n+1 prime.

The Seventeen or Bust project was conceived in March of 2002 by two college undergraduates. After some planning and a lot of programming, the first public client was released on April 1. Seventeen or Bust ceased operations in 2016. The project was administered by:
• Louis Helm, a computer engineer in Austin, Texas.
• David Norris, a software engineer in Urbana, Illinois.
• Michael Garrison, a Computer Science undergraduate at Eastern Michigan University in Ypsilanti, Michigan.

Starting in 2010, PrimeGrid partnered with Seventeen or Bust to work towards solving the Sierpinski Problem. After the demise of the original Seventeen or Bust project in 2016, PrimeGrid is continuing by itself to continue the Seventeen or Bust project in looking to solve the Sierpinski Problem.

As of December of 2017, PrimeGrid and Seventeen or Bust have eliminated twelve of those seventeen candidates. The project might now be styled "Five or Bust," but the original name will be kept for consistency. Current remaining k's:

21181, 22699, 24737, 55459, 67607

Last Prime found at PrimeGrid
10223*2^31172165+1 by Szabolcs Péter (SyP) on 31 October 2016. Official Announcement.
Previous to that the last SOB candidate was eliminated in 2007.

What is LLR?

The Lucas-Lehmer-Riesel (LLR) test is a primality test for numbers of the form N = k*2^n − 1, with 2^n > k. Also, LLR is a program developed by Jean Penne that can run the LLR-tests. It includes the Proth test to perform +1 tests and PRP to test non base 2 numbers. See also:

What is LLR2?

LLR2 is an improvement to the LLR application developed by our very own Pavel Atnashev and stream. It utilizes Gerbicz checks to enable the Fast DoubleCheck feature, which will nearly double the speed of PrimeGrid's progress on the projects it's applied to. For more information, see this forum post.
____________

Time flies like an arrow. Fruit flies like a banana.

Michael Goetz
Volunteer moderator

Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 3,564

Message 149257 - Posted: 8 Mar 2021 | 22:50:03 UTC

A head's up to anyone who was planning on using cloud computing resources from Digital Ocean, or, by extension, from TSC: Digital Ocean has changed their offerings, and it's no longer possible to get the 3-core AVX512 droplets we've been using in the past.

The basic droplets no longer support AVX512, and we have to use the premium droplets. While they're 20% more expensive, they're also newer CPUs and are worth the cost. However, they're not available in a 3-core configuration. You can still get basic 3-core droplets, but they won't have AVX512. There are 2 and 4 core premium droplets available, but at double the per-core cost of the 1 core droplets.

This is particularly poor timing coming right before the SoB challenge.
____________
My lucky number is 75898524288+1

freestman

Joined: 14 Jun 08
Posts: 30
ID: 24140
Credit: 253,866,835
RAC: 87,599

Message 149268 - Posted: 9 Mar 2021 | 6:26:48 UTC

This is really a CPU challenge project, my CPU can only complete one effective calculation in the schedule.
____________

Allen Paschke

Joined: 12 Nov 15
Posts: 26
ID: 428118
Credit: 38,541,389
RAC: 20,191

Message 149295 - Posted: 10 Mar 2021 | 0:57:14 UTC - in response to Message 149268.

I agree with freestman. Each of my 3 computers, each running a SOB task with all 4 threads, will complete one SOB task during the challenge. It will take each computer 6 - 7 days to complete a task.

If the goal of the challenge is to concentrate on SOB, how about more than 10 days?

xii5ku

Joined: 17 Dec 16
Posts: 79
ID: 476505
Credit: 1,124,080,059
RAC: 117,835

Message 149297 - Posted: 10 Mar 2021 | 6:19:21 UTC - in response to Message 149257.

Michael Goetz wrote:
A head's up to anyone who was planning on using cloud computing resources from Digital Ocean, [...] basic droplets no longer support AVX512, and we have to use the premium droplets.

Before you concern yourselves with per-core performance of your machines (virtual or not), you need to know how much last-level cache the application instance can grab. And if this does not amount to "a lot", then you would want to know how much main memory bandwidth the application instance can have for itself.

Allen Paschke wrote:
Each of my 3 computers, each running a SOB task with all 4 threads, will complete one SOB task during the challenge. It will take each computer 6 - 7 days to complete a task.

Check whether these computers have single-channel or dual-channel memory. If the former, consider to purchase another SO-DIMM. Maybe the KabyLake-U can then make it in 5 days (if the thermal budget allows it); I don't know about Sandy and Ivy Bridge-M.

freestman

Joined: 14 Jun 08
Posts: 30
ID: 24140
Credit: 253,866,835
RAC: 87,599

Message 149298 - Posted: 10 Mar 2021 | 7:03:29 UTC - in response to Message 149295.

Test results

____________

Chooka

Joined: 15 May 18
Posts: 267
ID: 1014486
Credit: 594,664,635
RAC: 2,311,640

Message 149299 - Posted: 10 Mar 2021 | 8:10:07 UTC

Sorry Freestman... could you please break down what I'm looking at with your results?

Lets take the 6700K for example.
Running all 8 threads, it takes 5.48 days to complete 8 tasks.
Running 4 cores, it takes on average around 4 days to complete.
I'm struggling to work out if it's better to run 8T or 4C based on your figures sorry. If I'm reading it right, you'll finish 8 tasks in 5.48 days vs 4 tasks in around 4 days. Is that right?

I didn't think running all threads was ever better than having H/T off? Either way, you'll only complete 8 tasks within the 10 day period.

Did I get that right?

____________

**Please boycott BOINC project Rakesearch as it's a Russian project! **

xii5ku

Joined: 17 Dec 16
Posts: 79
ID: 476505
Credit: 1,124,080,059
RAC: 117,835

Message 149310 - Posted: 10 Mar 2021 | 18:14:40 UTC - in response to Message 149299.

Chooka wrote:
Running all 8 threads, it takes 5.48 days to complete 8 tasks.

Chooka wrote:
I'm struggling to work out if it's better to run 8T or 4C based on your figures

My guess is that each row represents one test, and that a different, random workunit was used in each of these tests. (The numbers which were tested could have been put into this report to make it more useful. Or the credit granted for the result. Or even better: Both.)

In order to find out whether or not to use SMT with a given CPU, the SMT-on and SMT-off configurations need to be tested with the very same workunit.

If you can't (or don't want to) perform multiple tests with the same WU, then you (1.) need more tests, (2.) need to normalize the test results to the granted credit.

Bur
Volunteer tester

Joined: 25 Feb 20
Posts: 511
ID: 1241833
Credit: 408,530,855
RAC: 27,327

Message 149312 - Posted: 10 Mar 2021 | 18:44:26 UTC

Since SMT is a method to use unused CPU cycles, I doubt it makes crunching more effective in any scenario. Since there shouldn't be any unused cycles and SMT introduces an overhead, it should always be slower. Personally, I never had higher throughput using it.

You can get an idea by running Prime95 benchmarks with the FFT size of SoB.

What really does help btw is assigning cores to LLR tasks if SMT is anabled in the BIOS. That way the windows scheduler doesn't waste time moving the threads to logical cores and back. Single threaded tasks increased by about 15%, multithreaded still by a couple %.
____________
1281979 * 2^485014 + 1 is prime ... no further hits up to: n = 5,700,000

xii5ku

Joined: 17 Dec 16
Posts: 79
ID: 476505
Credit: 1,124,080,059
RAC: 117,835

Message 149319 - Posted: 10 Mar 2021 | 20:16:30 UTC - in response to Message 149312.

Bur wrote:
SMT [...] should always be slower.

With some LLR based sub-projects (maybe not llrSOB, I don't know) it gives slightly higher throughput on some processors.

Though at the few occasions at which I measured not only throughput but also power consumption, SMT gave lower throughput/Watt i.e. higher energy use per task.

Bur wrote:
the windows scheduler

I am lazy and use an OS with a sensible scheduler instead.

j.sheridan
Volunteer tester

Joined: 21 Mar 11
Posts: 816
ID: 91622
Credit: 2,686,055,085
RAC: 0

Message 149320 - Posted: 10 Mar 2021 | 20:25:20 UTC - in response to Message 149319.

Bur wrote:
the windows scheduler

I am lazy and use an OS with a sensible scheduler instead.

Letting Linux run wild is still slower than assigning affinity in windows (on Ryzen).

mackerel
Volunteer tester

Joined: 2 Oct 08
Posts: 2584
ID: 29980
Credit: 550,654,016
RAC: 7,286

Message 149328 - Posted: 10 Mar 2021 | 23:26:36 UTC - in response to Message 149312.

Since SMT is a method to use unused CPU cycles, I doubt it makes crunching more effective in any scenario. Since there shouldn't be any unused cycles and SMT introduces an overhead, it should always be slower. Personally, I never had higher throughput using it.

It's not about cycles, but processor resources. LLR and similar are highly effective at maxing out the FPU, and in that scenario HT/SMT doesn't provide any benefit. Sieves are very different and get a good boost from using HT/SMT.

What really does help btw is assigning cores to LLR tasks if SMT is anabled in the BIOS. That way the windows scheduler doesn't waste time moving the threads to logical cores and back. Single threaded tasks increased by about 15%, multithreaded still by a couple %.

Might depend on the system configuration, but I never saw more than 10% improvement from setting affinity for 1 task per core, and recently it seems to be much less than that although I'd still recommend affinity in that use case. Multi-thread never showed a difference for me.

freestman

Joined: 14 Jun 08
Posts: 30
ID: 24140
Credit: 253,866,835
RAC: 87,599

Message 149335 - Posted: 11 Mar 2021 | 4:57:50 UTC - in response to Message 149299.

Sorry Freestman... could you please break down what I'm looking at with your results?

Lets take the 6700K for example.
Running all 8 threads, it takes 5.48 days to complete 8 tasks.
Running 4 cores, it takes on average around 4 days to complete.
I'm struggling to work out if it's better to run 8T or 4C based on your figures sorry. If I'm reading it right, you'll finish 8 tasks in 5.48 days vs 4 tasks in around 4 days. Is that right?

I didn't think running all threads was ever better than having H/T off? Either way, you'll only complete 8 tasks within the 10 day period.

Did I get that right?

Due to insufficient testing for HT, I can't be sure if it helps. However, it can be seen that HT has little impact on this project.
____________

freestman

Joined: 14 Jun 08
Posts: 30
ID: 24140
Credit: 253,866,835
RAC: 87,599

Message 149336 - Posted: 11 Mar 2021 | 5:04:20 UTC - in response to Message 149299.

Sorry Freestman... could you please break down what I'm looking at with your results?

Lets take the 6700K for example.
Running all 8 threads, it takes 5.48 days to complete 8 tasks.
Running 4 cores, it takes on average around 4 days to complete.
I'm struggling to work out if it's better to run 8T or 4C based on your figures sorry. If I'm reading it right, you'll finish 8 tasks in 5.48 days vs 4 tasks in around 4 days. Is that right?

I didn't think running all threads was ever better than having H/T off? Either way, you'll only complete 8 tasks within the 10 day period.

Did I get that right?

Two computers with the same configuration have obvious differences in performance, perhaps due to differences in heat dissipation.

____________

composite
Volunteer tester

Joined: 16 Feb 10
Posts: 1022
ID: 55391
Credit: 888,368,094
RAC: 136,574

Message 149337 - Posted: 11 Mar 2021 | 5:40:28 UTC - in response to Message 149336.

Two computers with the same configuration have obvious differences in performance, perhaps due to differences in heat dissipation.

Check the CAS latency of the RAM in these computers to see if they are they same, especially if they are second-hand. A company selling refurbished/recycled desktops for \$10 is thinking about saleability, not performance tuning.

xii5ku

Joined: 17 Dec 16
Posts: 79
ID: 476505
Credit: 1,124,080,059
RAC: 117,835

Message 149338 - Posted: 11 Mar 2021 | 6:50:00 UTC - in response to Message 149310.

xii5ku wrote:
In order to find out whether or not to use SMT with a given CPU, the SMT-on and SMT-off configurations need to be tested with the very same workunit.

If you can't (or don't want to) perform multiple tests with the same WU, then you (1.) need more tests, (2.) need to normalize the test results to the granted credit.

Re, normalization to credit per result: I just looked at some llrSOB results of a Windows computer which is not my own. The PPD which I can calculate from that still has a high variability. Still too high to properly determine what actual difference SMT-on and SMT-off really make.

However, given the cache size and RAM performance requirements of llrSOB, it is pretty safe to say that SMT does not help at with llrSOB in particular.

(Personally, I am not guessing how to configure my computers optimally. I rely on reproducible and precise tests instead. I currently have two types of computers myself, and took all measurements which I need on one of these computer types within just a few hours last evening, using the same WU in all measurements.)

Ken_g6
Volunteer developer

Joined: 4 Jul 06
Posts: 929
ID: 3110
Credit: 236,594,205
RAC: 7,998

Message 149421 - Posted: 14 Mar 2021 | 15:36:13 UTC

So, how many people's computers started an hour early? I know mine did. Stupid daylight saving time.

JeppeSN

Joined: 5 Apr 14
Posts: 1727
ID: 306875
Credit: 41,399,492
RAC: 13,511

Message 149432 - Posted: 14 Mar 2021 | 17:40:16 UTC - in response to Message 149421.

So, how many people's computers started an hour early? I know mine did. Stupid daylight saving time.

If you forget to advance your clock in the spring (March), then usually you come too late.

However, if your clock adjusts itself automatically (March), and advances by one hour by itself, but you think it is still the old time (wrt. UTC), then you can come too early, as here.

(Note: NORTHERN: In USA/Canada (most places), this challenge is in Daylight Saving Time. In Iran, this starts in standard time and ends in daylight time. In Europe (where relevant), Israel/Palestine, Jordan, Syria; and Mexico (most places), this challenge is in standard (winter) time. SOUTHERN: Where relevant (Chile, Paraguay, Australia (some places), New Zealand), this challenge is in daylight time.)

/JeppeSN

Bur
Volunteer tester

Joined: 25 Feb 20
Posts: 511
ID: 1241833
Credit: 408,530,855
RAC: 27,327

Message 149433 - Posted: 14 Mar 2021 | 17:41:17 UTC - in response to Message 149421.

Starting early sucks, did you abort those tasks?
____________
1281979 * 2^485014 + 1 is prime ... no further hits up to: n = 5,700,000

dannyridel
Volunteer tester

Joined: 3 Feb 19
Posts: 966
ID: 1097922
Credit: 38,658,054
RAC: 182,972

Message 149454 - Posted: 15 Mar 2021 | 9:53:36 UTC - in response to Message 149421.

So, how many people's computers started an hour early? I know mine did. Stupid daylight saving time.

You can always look at PG graphs to get an idea of time :)
____________
My lucky number is 6219*2^3374198+1

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149463 - Posted: 15 Mar 2021 | 14:31:21 UTC

Challenge: Sier"pi"nski's Birthday
App: 13 (SoB-LLR)
Fast DC tasks are NOT included.
(As of 2021-03-15 14:30:30 UTC)

5746 tasks have been sent out. [CPU/GPU/anonymous_platform: 5746 (100%) / 0 (0%) / 0 (0%)]

Of those tasks that have been sent out:

486 (8%) were aborted. [486 (8%) / 0 (0%) / 0 (0%)]
15 (0%) came back with some kind of an error. [15 (0%) / 0 (0%) / 0 (0%)]
151 (3%) have returned a successful result. [151 (3%) / 0 (0%) / 0 (0%)]
5094 (89%) are still in progress. [5094 (89%) / 0 (0%) / 0 (0%)]

Of the tasks that have been returned successfully:

141 (93%) are pending validation. [141 (93%) / 0 (0%) / 0 (0%)]
10 (7%) have been successfully validated. [10 (7%) / 0 (0%) / 0 (0%)]
0 (0%) were invalid. [0 (0%) / 0 (0%) / 0 (0%)]
0 (0%) are inconclusive. [0 (0%) / 0 (0%) / 0 (0%)]

The current leading edge (i.e., latest work unit for which work has actually been sent out to a host) is n=34934623. The leading edge was at n=34489460 at the beginning of the challenge. Since the challenge started, the leading edge has advanced 1.29% as much as it had prior to the challenge!
____________

Time flies like an arrow. Fruit flies like a banana.

Michael Goetz
Volunteer moderator

Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 3,564

Message 149468 - Posted: 15 Mar 2021 | 17:23:19 UTC

If you're wondering about FFT sizes, these are the current FFTs in use by all SOB candidates that have been turned into workunits. That's a set that includes all in-progress tasks as well as about 2000 tasks that are ready to be sent out.

+-------+---------+-------+------------+------------+ | k | FFT (K) | count | MIN N | MAX N | +-------+---------+-------+------------+------------+ | 21181 | 2,880 | 4 | 33,664,988 | 33,855,332 | | 21181 | 3,072 | 2291 | 33,894,380 | 35,134,628 | | 22699 | 2,880 | 2 | 33,617,350 | 33,703,534 | | 22699 | 3,072 | 893 | 33,889,798 | 35,135,038 | | 24737 | 2,880 | 1 | 33,425,863 | 33,425,863 | | 24737 | 3,072 | 2474 | 33,749,647 | 35,134,423 | | 55459 | 3,072 | 19 | 33,432,274 | 33,965,218 | | 55459 | 3,200 | 3202 | 33,977,998 | 35,135,194 | | 67607 | 3,200 | 544 | 33,580,107 | 34,841,691 | | 67607 | 3,456 | 278 | 34,842,851 | 35,133,867 | +-------+---------+-------+------------+------------+

Looking forward, this second table includes all of the SoB candidates up to n=37M.

+-------+---------+-------+------------+------------+ | k | FFT (K) | count | MIN N | MAX N | +-------+---------+-------+------------+------------+ | 21181 | 2,880 | 4 | 33,664,988 | 33,855,332 | | 21181 | 3,072 | 5026 | 33,894,380 | 36,156,980 | | 21181 | 3,200 | 2306 | 36,158,180 | 36,999,572 | | 22699 | 2,880 | 2 | 33,617,350 | 33,703,534 | | 22699 | 3,072 | 1862 | 33,889,798 | 35,999,110 | | 22699 | 3,200 | 1138 | 36,000,118 | 36,999,838 | | 24737 | 2,880 | 1 | 33,425,863 | 33,425,863 | | 24737 | 3,072 | 4380 | 33,749,647 | 35,804,983 | | 24737 | 3,200 | 3386 | 35,805,103 | 36,999,871 | | 55459 | 3,072 | 19 | 33,432,274 | 33,965,218 | | 55459 | 3,200 | 3852 | 33,977,998 | 35,309,506 | | 55459 | 3,456 | 6418 | 35,310,358 | 36,999,526 | | 67607 | 3,200 | 544 | 33,580,107 | 34,841,691 | | 67607 | 3,456 | 2071 | 34,842,851 | 36,998,891 | +-------+---------+-------+------------+------------+

There's a few old 2880K FFT tasks in progress, but mostly everything is 3072K, 3200K, or 3456K.
____________
My lucky number is 75898524288+1

Bur
Volunteer tester

Joined: 25 Feb 20
Posts: 511
ID: 1241833
Credit: 408,530,855
RAC: 27,327

Message 149470 - Posted: 15 Mar 2021 | 21:17:27 UTC - in response to Message 149468.

Approximately what was the percentage of removed candidates for the 5 k's during sieving?

I'm asking to estimate from the number of completed WUs how quickly n increases over time.
____________
1281979 * 2^485014 + 1 is prime ... no further hits up to: n = 5,700,000

JeppeSN

Joined: 5 Apr 14
Posts: 1727
ID: 306875
Credit: 41,399,492
RAC: 13,511

Message 149472 - Posted: 15 Mar 2021 | 21:53:37 UTC - in response to Message 149470.

Approximately what was the percentage of removed candidates for the 5 k's during sieving?

I'm asking to estimate from the number of completed WUs how quickly n increases over time.

Not a (full) answer to your question, but Yves's calculator weight.js can write:
C(21181)[5e+05, 5e+06] = 0.1895 (57.6 s) C(22699)[5e+05, 5e+06] = 0.0777 (59.6 s) C(24737)[5e+05, 5e+06] = 0.1996 (66.7 s) C(55459)[5e+05, 5e+06] = 0.2659 (60.8 s) C(67607)[5e+05, 5e+06] = 0.0675 (60.8 s)

(these numbers are also in the PDF file). You can see from the number of seconds that I did this from a slow machine.

/JeppeSN

Michael Goetz
Volunteer moderator

Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 3,564

Message 149477 - Posted: 15 Mar 2021 | 23:37:53 UTC - in response to Message 149470.

Approximately what was the percentage of removed candidates for the 5 k's during sieving?

I'm asking to estimate from the number of completed WUs how quickly n increases over time.

After sieving, there were 11409 candidates 35M < n < 36M.

There's 5 k's, so there should be 5 million candidates in that range before sieving.

That's 0.23% after sieving, or, alternatively, sieving removed 99,77% of the candidates.

Also, note that P-1 factorization was also done on this sieve file, so those numbers reflect the combination of sieving and factorization.
____________
My lucky number is 75898524288+1

Bur
Volunteer tester

Joined: 25 Feb 20
Posts: 511
ID: 1241833
Credit: 408,530,855
RAC: 27,327

Message 149482 - Posted: 16 Mar 2021 | 7:36:19 UTC - in response to Message 149477.

Thanks!

Do you have the number of candidates for the remaining ranges? How far was SoB sieved/P-1 factorized?

And just out of curiosity, for which exponent size is P-1 factorization considered useful?
____________
1281979 * 2^485014 + 1 is prime ... no further hits up to: n = 5,700,000

Michael Goetz
Volunteer moderator

Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 3,564

Message 149484 - Posted: 16 Mar 2021 | 11:38:02 UTC - in response to Message 149482.

Do you have the number of candidates for the remaining ranges? How far was SoB sieved/P-1 factorized?

It's about the same for each range. The rate at which candidates are removed varies with P (the sieve depth) but does not vary with n.

Sieving and factorization goes up to 50M. It will be a long time until we need another sieve file.

And just out of curiosity, for which exponent size is P-1 factorization considered useful?

I think the rule of thumb is it's useful above n=35M, but I'm not really sure.

____________
My lucky number is 75898524288+1

Bur
Volunteer tester

Joined: 25 Feb 20
Posts: 511
ID: 1241833
Credit: 408,530,855
RAC: 27,327

Message 149485 - Posted: 16 Mar 2021 | 12:56:20 UTC - in response to Message 149484.

Interesting fact about the removal ratio. I noticed this in a sieve for n < 4M but wasn't sure if it remained that way for larger n.

On average 80 SoB tasks are completed each day or 16 per day and k. That is approximately 7000 n per day and k or 2.5M per year*k. Not taking into account that each task gets slower and slower...
____________
1281979 * 2^485014 + 1 is prime ... no further hits up to: n = 5,700,000

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149486 - Posted: 16 Mar 2021 | 13:34:00 UTC

Challenge: Sier"pi"nski's Birthday
App: 13 (SoB-LLR)
Fast DC tasks are NOT included.
(As of 2021-03-16 13:33:33 UTC)

9146 tasks have been sent out. [CPU/GPU/anonymous_platform: 9146 (100%) / 0 (0%) / 0 (0%)]

Of those tasks that have been sent out:

2696 (29%) were aborted. [2696 (29%) / 0 (0%) / 0 (0%)]
31 (0%) came back with some kind of an error. [31 (0%) / 0 (0%) / 0 (0%)]
723 (8%) have returned a successful result. [723 (8%) / 0 (0%) / 0 (0%)]
5696 (62%) are still in progress. [5696 (62%) / 0 (0%) / 0 (0%)]

Of the tasks that have been returned successfully:

555 (77%) are pending validation. [555 (77%) / 0 (0%) / 0 (0%)]
168 (23%) have been successfully validated. [168 (23%) / 0 (0%) / 0 (0%)]
0 (0%) were invalid. [0 (0%) / 0 (0%) / 0 (0%)]
0 (0%) are inconclusive. [0 (0%) / 0 (0%) / 0 (0%)]

The current leading edge (i.e., latest work unit for which work has actually been sent out to a host) is n=35091031. The leading edge was at n=34489460 at the beginning of the challenge. Since the challenge started, the leading edge has advanced 1.74% as much as it had prior to the challenge!
____________

Time flies like an arrow. Fruit flies like a banana.

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149507 - Posted: 17 Mar 2021 | 13:58:20 UTC

Day 3:

Challenge: Sier"pi"nski's Birthday
App: 13 (SoB-LLR)
Fast DC tasks are NOT included.
(As of 2021-03-17 13:57:51 UTC)

12281 tasks have been sent out. [CPU/GPU/anonymous_platform: 12281 (100%) / 0 (0%) / 0 (0%)]

Of those tasks that have been sent out:

3446 (28%) were aborted. [3446 (28%) / 0 (0%) / 0 (0%)]
236 (2%) came back with some kind of an error. [236 (2%) / 0 (0%) / 0 (0%)]
1616 (13%) have returned a successful result. [1616 (13%) / 0 (0%) / 0 (0%)]
6983 (57%) are still in progress. [6983 (57%) / 0 (0%) / 0 (0%)]

Of the tasks that have been returned successfully:

904 (56%) are pending validation. [904 (56%) / 0 (0%) / 0 (0%)]
712 (44%) have been successfully validated. [712 (44%) / 0 (0%) / 0 (0%)]
0 (0%) were invalid. [0 (0%) / 0 (0%) / 0 (0%)]
0 (0%) are inconclusive. [0 (0%) / 0 (0%) / 0 (0%)]

The current leading edge (i.e., latest work unit for which work has actually been sent out to a host) is n=35235031. The leading edge was at n=34489460 at the beginning of the challenge. Since the challenge started, the leading edge has advanced 2.16% as much as it had prior to the challenge!
____________

Time flies like an arrow. Fruit flies like a banana.

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149537 - Posted: 18 Mar 2021 | 21:51:55 UTC

Day 4:
Challenge: Sier"pi"nski's Birthday
App: 13 (SoB-LLR)
Fast DC tasks are NOT included.
(As of 2021-03-18 21:51:25 UTC)

14300 tasks have been sent out. [CPU/GPU/anonymous_platform: 14300 (100%) / 0 (0%) / 0 (0%)]

Of those tasks that have been sent out:

3695 (26%) were aborted. [3695 (26%) / 0 (0%) / 0 (0%)]
276 (2%) came back with some kind of an error. [276 (2%) / 0 (0%) / 0 (0%)]
2683 (19%) have returned a successful result. [2683 (19%) / 0 (0%) / 0 (0%)]
7646 (53%) are still in progress. [7646 (53%) / 0 (0%) / 0 (0%)]

Of the tasks that have been returned successfully:

828 (31%) are pending validation. [828 (31%) / 0 (0%) / 0 (0%)]
1855 (69%) have been successfully validated. [1855 (69%) / 0 (0%) / 0 (0%)]
0 (0%) were invalid. [0 (0%) / 0 (0%) / 0 (0%)]
0 (0%) are inconclusive. [0 (0%) / 0 (0%) / 0 (0%)]

The current leading edge (i.e., latest work unit for which work has actually been sent out to a host) is n=35372134. The leading edge was at n=34489460 at the beginning of the challenge. Since the challenge started, the leading edge has advanced 2.56% as much as it had prior to the challenge!
____________

Time flies like an arrow. Fruit flies like a banana.

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149548 - Posted: 19 Mar 2021 | 15:07:31 UTC

Day 5:
Challenge: Sier"pi"nski's Birthday
App: 13 (SoB-LLR)
Fast DC tasks are NOT included.
(As of 2021-03-19 15:07:02 UTC)

16733 tasks have been sent out. [CPU/GPU/anonymous_platform: 16733 (100%) / 0 (0%) / 0 (0%)]

Of those tasks that have been sent out:

4612 (28%) were aborted. [4612 (28%) / 0 (0%) / 0 (0%)]
293 (2%) came back with some kind of an error. [293 (2%) / 0 (0%) / 0 (0%)]
3208 (19%) have returned a successful result. [3208 (19%) / 0 (0%) / 0 (0%)]
8620 (52%) are still in progress. [8620 (52%) / 0 (0%) / 0 (0%)]

Of the tasks that have been returned successfully:

859 (27%) are pending validation. [859 (27%) / 0 (0%) / 0 (0%)]
2349 (73%) have been successfully validated. [2349 (73%) / 0 (0%) / 0 (0%)]
0 (0%) were invalid. [0 (0%) / 0 (0%) / 0 (0%)]
0 (0%) are inconclusive. [0 (0%) / 0 (0%) / 0 (0%)]

The current leading edge (i.e., latest work unit for which work has actually been sent out to a host) is n=35569702. The leading edge was at n=34489460 at the beginning of the challenge. Since the challenge started, the leading edge has advanced 3.13% as much as it had prior to the challenge!
____________

Time flies like an arrow. Fruit flies like a banana.

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149559 - Posted: 20 Mar 2021 | 14:50:59 UTC

Day 6:

Challenge: Sier"pi"nski's Birthday
App: 13 (SoB-LLR)
Fast DC tasks are NOT included.
(As of 2021-03-20 14:50:25 UTC)

18011 tasks have been sent out. [CPU/GPU/anonymous_platform: 18011 (100%) / 0 (0%) / 0 (0%)]

Of those tasks that have been sent out:

4899 (27%) were aborted. [4899 (27%) / 0 (0%) / 0 (0%)]
315 (2%) came back with some kind of an error. [315 (2%) / 0 (0%) / 0 (0%)]
4311 (24%) have returned a successful result. [4311 (24%) / 0 (0%) / 0 (0%)]
8486 (47%) are still in progress. [8486 (47%) / 0 (0%) / 0 (0%)]

Of the tasks that have been returned successfully:

1264 (29%) are pending validation. [1264 (29%) / 0 (0%) / 0 (0%)]
3047 (71%) have been successfully validated. [3047 (71%) / 0 (0%) / 0 (0%)]
0 (0%) were invalid. [0 (0%) / 0 (0%) / 0 (0%)]
0 (0%) are inconclusive. [0 (0%) / 0 (0%) / 0 (0%)]

The current leading edge (i.e., latest work unit for which work has actually been sent out to a host) is n=35595874. The leading edge was at n=34489460 at the beginning of the challenge. Since the challenge started, the leading edge has advanced 3.21% as much as it had prior to the challenge!
____________

Time flies like an arrow. Fruit flies like a banana.

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149577 - Posted: 21 Mar 2021 | 13:22:00 UTC

Day 7:
Challenge: Sier"pi"nski's Birthday
App: 13 (SoB-LLR)
Fast DC tasks are NOT included.
(As of 2021-03-21 13:21:41 UTC)

20526 tasks have been sent out. [CPU/GPU/anonymous_platform: 20526 (100%) / 0 (0%) / 0 (0%)]

Of those tasks that have been sent out:

7934 (39%) were aborted. [7934 (39%) / 0 (0%) / 0 (0%)]
345 (2%) came back with some kind of an error. [345 (2%) / 0 (0%) / 0 (0%)]
5334 (26%) have returned a successful result. [5334 (26%) / 0 (0%) / 0 (0%)]
6913 (34%) are still in progress. [6913 (34%) / 0 (0%) / 0 (0%)]

Of the tasks that have been returned successfully:

1464 (27%) are pending validation. [1464 (27%) / 0 (0%) / 0 (0%)]
3870 (73%) have been successfully validated. [3870 (73%) / 0 (0%) / 0 (0%)]
1 (0%) were invalid. [1 (0%) / 0 (0%) / 0 (0%)]
0 (0%) are inconclusive. [0 (0%) / 0 (0%) / 0 (0%)]

The current leading edge (i.e., latest work unit for which work has actually been sent out to a host) is n=35629011. The leading edge was at n=34489460 at the beginning of the challenge. Since the challenge started, the leading edge has advanced 3.30% as much as it had prior to the challenge!
____________

Time flies like an arrow. Fruit flies like a banana.

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149604 - Posted: 22 Mar 2021 | 14:18:18 UTC

Day 8:

Challenge: Sier"pi"nski's Birthday
App: 13 (SoB-LLR)
Fast DC tasks are NOT included.
(As of 2021-03-22 14:17:59 UTC)

22238 tasks have been sent out. [CPU/GPU/anonymous_platform: 22238 (100%) / 0 (0%) / 0 (0%)]

Of those tasks that have been sent out:

9701 (44%) were aborted. [9701 (44%) / 0 (0%) / 0 (0%)]
365 (2%) came back with some kind of an error. [365 (2%) / 0 (0%) / 0 (0%)]
6382 (29%) have returned a successful result. [6382 (29%) / 0 (0%) / 0 (0%)]
5790 (26%) are still in progress. [5790 (26%) / 0 (0%) / 0 (0%)]

Of the tasks that have been returned successfully:

1529 (24%) are pending validation. [1529 (24%) / 0 (0%) / 0 (0%)]
4853 (76%) have been successfully validated. [4853 (76%) / 0 (0%) / 0 (0%)]
1 (0%) were invalid. [1 (0%) / 0 (0%) / 0 (0%)]
0 (0%) are inconclusive. [0 (0%) / 0 (0%) / 0 (0%)]

The current leading edge (i.e., latest work unit for which work has actually been sent out to a host) is n=35629011. The leading edge was at n=34489460 at the beginning of the challenge. Since the challenge started, the leading edge has advanced 3.30% as much as it had prior to the challenge!
____________

Time flies like an arrow. Fruit flies like a banana.

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149629 - Posted: 23 Mar 2021 | 13:34:42 UTC

Day 9:
Challenge: Sier"pi"nski's Birthday
App: 13 (SoB-LLR)
Fast DC tasks are NOT included.
(As of 2021-03-23 13:34:24 UTC)

24768 tasks have been sent out. [CPU/GPU/anonymous_platform: 24768 (100%) / 0 (0%) / 0 (0%)]

Of those tasks that have been sent out:

11697 (47%) were aborted. [11697 (47%) / 0 (0%) / 0 (0%)]
385 (2%) came back with some kind of an error. [385 (2%) / 0 (0%) / 0 (0%)]
7360 (30%) have returned a successful result. [7360 (30%) / 0 (0%) / 0 (0%)]
5326 (22%) are still in progress. [5326 (22%) / 0 (0%) / 0 (0%)]

Of the tasks that have been returned successfully:

1789 (24%) are pending validation. [1789 (24%) / 0 (0%) / 0 (0%)]
5571 (76%) have been successfully validated. [5571 (76%) / 0 (0%) / 0 (0%)]
2 (0%) were invalid. [2 (0%) / 0 (0%) / 0 (0%)]
0 (0%) are inconclusive. [0 (0%) / 0 (0%) / 0 (0%)]

The current leading edge (i.e., latest work unit for which work has actually been sent out to a host) is n=35629011. The leading edge was at n=34489460 at the beginning of the challenge. Since the challenge started, the leading edge has advanced 3.30% as much as it had prior to the challenge!
____________

Time flies like an arrow. Fruit flies like a banana.

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149649 - Posted: 24 Mar 2021 | 13:32:17 UTC

Some friendly reminders... :)

At the Conclusion of the Challenge

When the challenge completes, we would prefer users "moving on" to finish those tasks they have downloaded, if not then please ABORT the WU's (and then UPDATE the PrimeGrid project) instead of DETACHING, RESETTING, or PAUSING.

ABORTING WU's allows them to be recycled immediately; thus a much faster "clean up" to the end of a Challenge. DETACHING, RESETTING, and PAUSING WU's causes them to remain in limbo until they EXPIRE. Therefore, we must wait until WU's expire to send them out to be completed.

Likewise, if you're shutting down the computer for an extended period of time, or deleting the VM (Virtual Machine), please ABORT all remaining tasks first. Also, be aware that merely shutting off a cloud server doesn't stop the billing. You have to destroy/delete the server if you don't want to continue to be charged for it.

Thank you!
____________

Time flies like an arrow. Fruit flies like a banana.

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149651 - Posted: 24 Mar 2021 | 16:24:07 UTC

And we're done! Final stats below. Cleanup is starting now, currently 1260 tasks outstanding; 1006 affecting individual (209) scoring positions; 768 affecting team (51) scoring positions.

Challenge: Sier"pi"nski's Birthday
App: 13 (SoB-LLR)
Fast DC tasks are NOT included.
(As of 2021-03-24 13:32:38 UTC)

26743 tasks have been sent out. [CPU/GPU/anonymous_platform: 26743 (100%) / 0 (0%) / 0 (0%)]

Of those tasks that have been sent out:

13362 (50%) were aborted. [13362 (50%) / 0 (0%) / 0 (0%)]
401 (1%) came back with some kind of an error. [401 (1%) / 0 (0%) / 0 (0%)]
8420 (31%) have returned a successful result. [8420 (31%) / 0 (0%) / 0 (0%)]
4454 (17%) are still in progress. [4454 (17%) / 0 (0%) / 0 (0%)]

Of the tasks that have been returned successfully:

1465 (17%) are pending validation. [1465 (17%) / 0 (0%) / 0 (0%)]
6955 (83%) have been successfully validated. [6955 (83%) / 0 (0%) / 0 (0%)]
2 (0%) were invalid. [2 (0%) / 0 (0%) / 0 (0%)]
0 (0%) are inconclusive. [0 (0%) / 0 (0%) / 0 (0%)]

The current leading edge (i.e., latest work unit for which work has actually been sent out to a host) is n=35629011. The leading edge was at n=34489460 at the beginning of the challenge. Since the challenge started, the leading edge has advanced 3.30% as much as it had prior to the challenge!
____________

Time flies like an arrow. Fruit flies like a banana.

WezH

Joined: 9 Jun 11
Posts: 95
ID: 101605
Credit: 461,352,862
RAC: 1,071,336

Message 149653 - Posted: 24 Mar 2021 | 17:06:13 UTC - in response to Message 149651.

Cleanup is starting now, currently 1260 tasks outstanding; 1006 affecting individual (209) scoring positions; 768 affecting team (51) scoring positions.

Or did I misunderstand "... received BEFORE 24th March 2021 11:59 UTC will be considered for credit", tasks returned after that will somehow counted to score?

Ravi Fernando
Volunteer tester
Project scientist

Joined: 21 Mar 19
Posts: 206
ID: 1108183
Credit: 11,995,107
RAC: 5,668

Message 149654 - Posted: 24 Mar 2021 | 17:31:22 UTC - in response to Message 149653.

Correct.

Chooka

Joined: 15 May 18
Posts: 267
ID: 1014486
Credit: 594,664,635
RAC: 2,311,640

Message 149659 - Posted: 25 Mar 2021 | 8:00:52 UTC

Thanks for organising the challenge Michael.
Its amazing to see how well one can do when not up against a lot of people renting servers to crunch with ie - the TdP) :D (or perhaps those users just couldn't be bothered with SoB)

Till the WW challenge! See you then.

____________

**Please boycott BOINC project Rakesearch as it's a Russian project! **

Scott Brown
Volunteer moderator
Volunteer tester
Project scientist

Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,607,819,909
RAC: 11,929,533

Message 149664 - Posted: 25 Mar 2021 | 18:30:53 UTC

Doing a little post challenge analysis of my results and found a comparison that was an interesting demonstration of the value of older hardware with big cache relative to newer hardware.

The two contenders:

1) i7-9700, 8-core (no HT), 3.0GHz, 12MB L3 cache, DDR4 2666 dual channel memory, AVX2/FMA3 (Coffee Lake). about \$270 US on NewEgg

2) Xeon E5-2687W, 8-core (HT on, but running 50% CPU), 3.1GHz, 20MB L3 cache, DDR3 1600 quad channel memory, AVX (Sandy Bridge). about \$100 US on Ebay

SoB FFT = 3M
i7-9700 run is about 135k seconds (or 37.5 hours)

SoB FFT = 3456k
i7-9700 run is about 164k seconds (or 45.5 hours)

With nearly equal clocks on these two CPUs, it seems clear that for large work like SoB, bigger cache and more memory channels trumps the newer instruction set and faster memory.

Chooka

Joined: 15 May 18
Posts: 267
ID: 1014486
Credit: 594,664,635
RAC: 2,311,640

Message 149665 - Posted: 25 Mar 2021 | 18:57:15 UTC

Interesting!
Thanks for sharing Scott.

I really noticed during the challenge just how much quicker an Intel PC with AVX-512 was compared to say a Ryzen 3950X.

A 9960X (16 cores) did SoB w/u's in around... 30-33K sec per w/u vs a 3950X (16 cores) which was around 70 - 80K per w/u.

____________

**Please boycott BOINC project Rakesearch as it's a Russian project! **

Nick

Joined: 11 Jul 11
Posts: 1955
ID: 105020
Credit: 5,822,855,097
RAC: 33,404,624

Message 149667 - Posted: 25 Mar 2021 | 20:30:20 UTC - in response to Message 149665.

Interesting!
Thanks for sharing Scott.

I really noticed during the challenge just how much quicker an Intel PC with AVX-512 was compared to say a Ryzen 3950X.

A 9960X (16 cores) did SoB w/u's in around... 30-33K sec per w/u vs a 3950X (16 cores) which was around 70 - 80K per w/u.

I think I'm the only person with 9960Xs.
What is interesting is that some tasks ran much slower.
All tasks with an FFT of 3360K were fast (33K sec)
A fast task with FFT of 3000K was 29K sec.

Even my 9980XE was nearly as fast as a 9960X for fast tasks and it has shown that it can nearly keep up - so now I know that it slacks off. Up until running this challenge, in nearly every project, the 9960Xs were faster than the 9980XE by about 10-15% and run about 10 degrees C cooler. Go figure?

I know an AMD 5950X that was doing 2 x 8 thread tasks in 54K sec which was with FFT of 3000K
And 65K sec for FFT of 3456K - these are faster than my 9960Xs by about 7% for the FFT of 3000K

I learnt on the last day that I could download many tasks and quickly work out which ones were going to run as well as possible - aborting the rest.

mackerel
Volunteer tester

Joined: 2 Oct 08
Posts: 2584
ID: 29980
Credit: 550,654,016
RAC: 7,286

Message 149668 - Posted: 25 Mar 2021 | 23:40:51 UTC - in response to Message 149667.

Zen 3 = 32MB/CCX so with correct affinity set to run one task on each, work isn't limited by ram speed
Zen 2 = 16MB/CCX - never tested it in depth while I owned some but it was not a good arrangement for large tasks like these. If you can't keep a task within a single CCX then you cross CCX and the scaling is not good.

Most others without enough unified cache will be limited by ram speed. I never saw cache amount matters beyond enough or not enough e.g. Skylake 6600k (6MB cache) vs 6700k (8MB cache) made no difference if the tasks required more than that. However, I think that was before multi-thread was common and I don't remember if I specifically revisited that. I don't have the 6600k any more although I might be able to do something with an i3...

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149695 - Posted: 26 Mar 2021 | 22:44:46 UTC

Cleanup Status:
Mar 25: 512 tasks outstanding; 318 affecting individual (138) scoring positions; 127 affecting team (30) scoring positions.
Mar 26: 261 tasks outstanding; 124 affecting individual (80) scoring positions; 48 affecting team (17) scoring positions.
Mar 27: 122 tasks outstanding; 61 affecting individual (47) scoring positions; 10 affecting team (8) scoring positions.
Mar 28: 60 tasks outstanding; 30 affecting individual (24) scoring positions; 6 affecting team (6) scoring positions.
Mar 29: 28 tasks outstanding; 13 affecting individual (11) scoring positions; 2 affecting team (2) scoring positions.
____________

Time flies like an arrow. Fruit flies like a banana.

Chooka

Joined: 15 May 18
Posts: 267
ID: 1014486
Credit: 594,664,635
RAC: 2,311,640

Message 149745 - Posted: 30 Mar 2021 | 5:28:27 UTC - in response to Message 149667.

Interesting!
Thanks for sharing Scott.

I really noticed during the challenge just how much quicker an Intel PC with AVX-512 was compared to say a Ryzen 3950X.

A 9960X (16 cores) did SoB w/u's in around... 30-33K sec per w/u vs a 3950X (16 cores) which was around 70 - 80K per w/u.

I think I'm the only person with 9960Xs.
What is interesting is that some tasks ran much slower.
All tasks with an FFT of 3360K were fast (33K sec)
A fast task with FFT of 3000K was 29K sec.

Even my 9980XE was nearly as fast as a 9960X for fast tasks and it has shown that it can nearly keep up - so now I know that it slacks off. Up until running this challenge, in nearly every project, the 9960Xs were faster than the 9980XE by about 10-15% and run about 10 degrees C cooler. Go figure?

I know an AMD 5950X that was doing 2 x 8 thread tasks in 54K sec which was with FFT of 3000K
And 65K sec for FFT of 3456K - these are faster than my 9960Xs by about 7% for the FFT of 3000K

I learnt on the last day that I could download many tasks and quickly work out which ones were going to run as well as possible - aborting the rest.

Hi Nick,

Yes it was you I was referring to! :D
Your PC's with AVX-512 are really great for PG. I assume that's why you bought them? They are WAY faster than my 3950X
____________

**Please boycott BOINC project Rakesearch as it's a Russian project! **

j.sheridan
Volunteer tester

Joined: 21 Mar 11
Posts: 816
ID: 91622
Credit: 2,686,055,085
RAC: 0

Message 149750 - Posted: 30 Mar 2021 | 8:40:06 UTC - in response to Message 149745.

Not as quick as a 5950x though:-)

Nick

Joined: 11 Jul 11
Posts: 1955
ID: 105020
Credit: 5,822,855,097
RAC: 33,404,624

Message 149766 - Posted: 31 Mar 2021 | 8:30:49 UTC - in response to Message 149750.

Not as quick as a 5950x though:-)

The 5950x is faster with SOB (~7%) than a 9960X and most likely faster again with PSP.
Running 2 x CUL or 2 x WOO in cache is borderline with a 9960X - so I'd say the 9960X would be slower with those.
I don't know if GCW has become too large to run 2 x tasks in cache on a 9960X.
We'll have to see about smaller FFT projects.... :)

Nick

Joined: 11 Jul 11
Posts: 1955
ID: 105020
Credit: 5,822,855,097
RAC: 33,404,624

Message 149767 - Posted: 31 Mar 2021 | 8:55:35 UTC - in response to Message 149745.

Hi Nick,

Yes it was you I was referring to! :D
Your PC's with AVX-512 are really great for PG. I assume that's why you bought them? They are WAY faster than my 3950X

Guilty! I bought my 2nd computer - a 9960X - entirely for primegrid.
Initially I wasn't able to use AVX512 and comparing computer throughput, it ran not much better than the 8 core 9900K.
I remember that to get best throughput, it was better to use around 8 to 10 cores.
That all changed when AVX512 was introduced into LLR.

I initially chose the 9960X because I was considering a 9920X - I thought maybe the 9940X, but settled on the 9960X because of the extra cache.
But it left me wondering if I should have gone all out and got the 9980XE. So I got one.
It underperformed the 9960X, core for core by 15% and was hotter.

Then I got really silly and bought another 9960X confirming that it is (was) a better choice.
I say was because there are 2 or 3 projects that now test it's ability to run 2 tasks in cache at once, whereas the 9980XE should have an advantage with those 2 or 3 projects over the 9960X.
If I was to buy another intel, I would buy another 9960X (or newer AVX512 16 core if it exists). I am not going to buy any more hardware - I have already gone too far.
My 4th computer is ready to go for a real graphics card, but I am not going to go there.

Michael Millerick
Volunteer tester

Joined: 4 Feb 09
Posts: 855
ID: 35074
Credit: 582,219,374
RAC: 1,371,198

Message 149773 - Posted: 31 Mar 2021 | 14:21:59 UTC

How do you cool the 9960X and what temps do you get?
____________

Nick

Joined: 11 Jul 11
Posts: 1955
ID: 105020
Credit: 5,822,855,097
RAC: 33,404,624

Message 149775 - Posted: 31 Mar 2021 | 15:48:24 UTC - in response to Message 149773.

How do you cool the 9960X and what temps do you get?

On the new 9960X, it is running at 50 degrees C - it hasn't gone above 55 when the weather is hot.
This is with a Corsair iCUE H150i RGB PRO XT - 3 fans water cooled.

The older 9960X now has some issues and is running in the high 60s - even though not too long ago it was running nearly as cool as the new one.
The issues are that it turns out I have a Corsair iCUE H100i RGB PRO XT - 2 fans - even though I paid for the larger model. Another issue is the MB is a bit screwy and the mosfets are overheating. And a fan on the GPU is knackered. (I'm waiting for the right moment to make a warranty claim for the GPU and then address why I have the wrong cooling)

So right now, running 4 x SR5 on each 9960X and running only 2 x SR5 on the 9980XE because if I run more it goes into the high 70s for 3 tasks and into the 80s for 4 tasks. It usually doesn't run this hot, but it will be warmer than the old 9960X (which usually sits in the low 60s). Note it isn't the number of cores running SR5 - it is the number of tasks.

Vitaly

Joined: 24 Sep 12
Posts: 68
ID: 173343
Credit: 13,785,129
RAC: 0

Message 149815 - Posted: 2 Apr 2021 | 17:59:15 UTC - in response to Message 149775.

Daniel Broomfield

Joined: 14 Jun 09
Posts: 61
ID: 41878
Credit: 1,571,992
RAC: 51

Message 149819 - Posted: 3 Apr 2021 | 7:04:51 UTC - in response to Message 149815.

Yes, it finished on 24th March. The next challenge begins on 11th April, for the Wall-Sun-Sun / Wieferich Prime Search.
____________
Proud member of Aggie The Pew!

7881*2^327265+1 is prime!
9411*2^367623+1 is prime!

JeppeSN

Joined: 5 Apr 14
Posts: 1727
ID: 306875
Credit: 41,399,492
RAC: 13,511

Message 149820 - Posted: 3 Apr 2021 | 7:40:31 UTC - in response to Message 149815.

See the dates on PrimeGrid's Challenge Series. /JeppeSN

dannyridel
Volunteer tester

Joined: 3 Feb 19
Posts: 966
ID: 1097922
Credit: 38,658,054
RAC: 182,972

Message 149822 - Posted: 3 Apr 2021 | 9:24:59 UTC - in response to Message 149815.

I think you mean whether the results are final, in that case, no, since there are still several outstanding tasks.
____________
My lucky number is 6219*2^3374198+1

Chooka

Joined: 15 May 18
Posts: 267
ID: 1014486
Credit: 594,664,635
RAC: 2,311,640

Message 149825 - Posted: 3 Apr 2021 | 12:20:54 UTC - in response to Message 149767.

Hi Nick,

Yes it was you I was referring to! :D
Your PC's with AVX-512 are really great for PG. I assume that's why you bought them? They are WAY faster than my 3950X

Guilty! I bought my 2nd computer - a 9960X - entirely for primegrid.
Initially I wasn't able to use AVX512 and comparing computer throughput, it ran not much better than the 8 core 9900K.
I remember that to get best throughput, it was better to use around 8 to 10 cores.
That all changed when AVX512 was introduced into LLR.

I initially chose the 9960X because I was considering a 9920X - I thought maybe the 9940X, but settled on the 9960X because of the extra cache.
But it left me wondering if I should have gone all out and got the 9980XE. So I got one.
It underperformed the 9960X, core for core by 15% and was hotter.

Then I got really silly and bought another 9960X confirming that it is (was) a better choice.
I say was because there are 2 or 3 projects that now test it's ability to run 2 tasks in cache at once, whereas the 9980XE should have an advantage with those 2 or 3 projects over the 9960X.
If I was to buy another intel, I would buy another 9960X (or newer AVX512 16 core if it exists). I am not going to buy any more hardware - I have already gone too far.
My 4th computer is ready to go for a real graphics card, but I am not going to go there.

Hahaha there you go! :)

I hear you on the hardware front. Enough (is never enough) but it will have to do when it comes to crunching pc's.

It will be interesting to see if the rumour is true about Zen 4 supporting AVX-512.
Could be a worthy upgrade. Especially if it's on 5nm. (code name GENOA)

Start saving. :D

https://www.techpowerup.com/279129/amd-zen-4-microarchitecture-to-support-avx-512#:~:text=The%20next%2Dgeneration%20%22Zen%204,web%20on%20the%20ChipHell%20forums.
____________

**Please boycott BOINC project Rakesearch as it's a Russian project! **

Michael Gutierrez
Volunteer moderator
Project scientist

Joined: 21 Mar 17
Posts: 278
ID: 764476
Credit: 46,020,786
RAC: 6

Message 149827 - Posted: 3 Apr 2021 | 15:42:45 UTC

The results are final!

During these 10 days, we completed more than 8 thousand tasks. That may not sound like a lot compared to other challenges, but the size of SOB tasks means that's equivalent to more than 550 Million credits. Unfortunately, we didn't find any new primes, but our overall output was still more than 5 times SOB's average -- or about 10 tng's!

108 teams and 476 individuals participated in the Challenge.

Top Three Individuals:
1. Pavel Atnashev
2. tng
3. vaughan

Top Three Teams:
1. Czech National Team
2. Ural Federal University
3. SETI.Germany

The next Challenge is Yuri's Night Challenge, which starts April 11. We hope to see you there! Thank you again to everyone. We appreciate your participation and hope you had fun!
____________

Time flies like an arrow. Fruit flies like a banana.

dannyridel
Volunteer tester

Joined: 3 Feb 19
Posts: 966
ID: 1097922
Credit: 38,658,054
RAC: 182,972

Message 149838 - Posted: 4 Apr 2021 | 0:52:15 UTC - in response to Message 149827.

During these 10 days, we completed more than 8 thousand tasks. That may not sound like a lot compared to other challenges, but the size of SOB tasks means that's equivalent to more than 550 Million credits. Unfortunately, we didn't find any new primes, but our overall output was still more than 5 times SOB's average -- or about 10 tng's!

108 teams and 476 individuals participated in the Challenge.

So one tng is around 47.6 average users? ;) thats impressive!
____________
My lucky number is 6219*2^3374198+1

composite
Volunteer tester

Joined: 16 Feb 10
Posts: 1022
ID: 55391
Credit: 888,368,094
RAC: 136,574

Message 149858 - Posted: 5 Apr 2021 | 5:32:15 UTC - in response to Message 149838.

During these 10 days, we completed more than 8 thousand tasks. That may not sound like a lot compared to other challenges, but the size of SOB tasks means that's equivalent to more than 550 Million credits. Unfortunately, we didn't find any new primes, but our overall output was still more than 5 times SOB's average -- or about 10 tng's!

108 teams and 476 individuals participated in the Challenge.

So one tng is around 47.6 average users? ;) thats impressive!

I read that as 1 tng = 1/2 typical total daily SOB throughput. Just as impressive!

Message boards : Number crunching : Sier"pi"nski's Birthday Challenge