Posts by Sergei Chernykh

1) Message boards : Number crunching : BOINC Pentathlon (Message 1417)
Posted 13 days ago by Sergei Chernykh
Post:
It looks like the validation queue will be cleared by the end of today.
2) Message boards : Number crunching : BOINC Pentathlon (Message 1412)
Posted 14 days ago by Sergei Chernykh
Post:
I've checked RAID status with mdadm and it's all fine. It seems to be just a high load because of too short WUs.
3) Message boards : Number crunching : BOINC Pentathlon (Message 1410)
Posted 14 days ago by Sergei Chernykh
Post:
Nothing else but validator and scheduler is turned on now, but HDD just can't keep up. It didn't happen in the previous two years, I don't know why it's happening now.
4) Message boards : Number crunching : BOINC Pentathlon (Message 1406)
Posted 14 days ago by Sergei Chernykh
Post:
It's a 6-core dedicated machine with 64 GB RAM, but the problem is the slow spinning HDD there, sorry! I turned off all non-essential background services, hope it will help.
5) Message boards : Bug tracker : WU'S taking longer (Message 1397)
Posted 19 days ago by Sergei Chernykh
Post:
WUs are not all the same due to the nature of the search algorithm. While task generator tries to split the work into equal chunks, it's impossible to predict how much time exactly each WU will take.
6) Message boards : Number crunching : BOINC Pentathlon (Message 1391)
Posted 20 days ago by Sergei Chernykh
Post:
The server is ready for this.
7) Message boards : Number crunching : Um, Got a Couple of Whoppers? (Message 1379)
Posted 14 Apr 2020 by Sergei Chernykh
Post:
Yes, this was unusually slow. It's probably related to the nature of the search - some ranges can't be filtered well enough, so more computations per number are needed.
8) Message boards : Number crunching : Um, Got a Couple of Whoppers? (Message 1377)
Posted 13 Apr 2020 by Sergei Chernykh
Post:
Probably because you have RTX 2070 and the other guy has old "Radeon R9 200 Series".
9) Message boards : News : The search up to 10^21 (Message 1371)
Posted 31 Mar 2020 by Sergei Chernykh
Post:
It's a bit too early to talk about this, but here's the search layout:

Phase 1: all odd pairs (m, n) with largest prime factor of m < 1011
Phase 2: all even pairs (m, n) with largest prime factor of m < 1011
Phase 3: all remaining pairs (m, n) with largest prime factor of m > 1011

Phase 3 will take much longer than 1 and 2 combined.
10) Message boards : Number crunching : Um, Got a Couple of Whoppers? (Message 1365)
Posted 24 Mar 2020 by Sergei Chernykh
Post:
This is strange. Your wingmen didn't seem to have any problems with these 2 WUs. Both were running GPU version though.
11) Message boards : Bug tracker : Supermassive WUs all ending in computation errors (Message 1360)
Posted 4 Mar 2020 by Sergei Chernykh
Post:
I just added the second batch and changed status to "part 2", but it actually displays both batches now.
12) Message boards : Bug tracker : Supermassive WUs all ending in computation errors (Message 1358)
Posted 4 Mar 2020 by Sergei Chernykh
Post:
The new batch hasn't started yet. It will have factorizations starting with
2N*...
13) Message boards : Bug tracker : Supermassive WUs all ending in computation errors (Message 1353)
Posted 16 Feb 2020 by Sergei Chernykh
Post:
This was a bug in the task generator, next batch of tasks (2N ranges) will never have tasks with more than ~500 billion numbers to check.
14) Message boards : Bug tracker : large fail rate (Message 1350)
Posted 13 Feb 2020 by Sergei Chernykh
Post:
For some reason your GPU runs out of memory. Do you run more than 1 task per GPU?
15) Message boards : Bug tracker : Factorization '0' is incorrect: 0 is not a prime! (Message 1346)
Posted 5 Feb 2020 by Sergei Chernykh
Post:
This error happens when saved progress ("amicable_checkpoint" file) contains incorrect data or was corrupted somehow. If you get this error again, please check "amicable_checkpoint" in the app directory and post it here.
16) Message boards : Random stuff : Amicable numbers divisible neither 2 nor 3. (Message 1338)
Posted 7 Jan 2020 by Sergei Chernykh
Post:
Here's the list: https://sech.me/ap/coprime_to_6.txt. It's very easy to find such pairs with the current method of search, I just went through all factorization ranges starting with 5N for up to 24-digit numbers. One PC was enough for it.
17) Message boards : Random stuff : Amicable Pairs Discovered (Message 1334)
Posted 29 Dec 2019 by Sergei Chernykh
Post:
There's no fancy interface for this, but all new pairs are logged: https://sech.me/ap/log/2019/2019-12-29/ - start from there and surf between different dates.
18) Message boards : Random stuff : Amicable numbers divisible neither 2 nor 3. (Message 1330)
Posted 20 Dec 2019 by Sergei Chernykh
Post:
Yes, and there is a lot of them. Smallest examples:
X44 Walker&Einstein 2001
5480828320492525=5^2*7^2*11*13*19*31*17*23*103*1319
5786392931193875=5^3*7*11*13*19*31*37*43*61*809

X43 Chernykh 2015
445953248528881275=3^2*5^2*7*13*19*37*43*73*439*22483
659008669204392325=5^2*7*13*19*37*73*571*1693*5839
19) Message boards : Number crunching : GPU version requirements for the search up to 10^20 (Message 1326)
Posted 13 Dec 2019 by Sergei Chernykh
Post:
Because GPU version has to store all prime numbers < 1011 and it can't use compact bit array as CPU version does.
20) Message boards : Bug tracker : badly WU? (Message 1317)
Posted 5 Dec 2019 by Sergei Chernykh
Post:
Yes, looks like it. I'll have to split it into smaller WUs.
21) Message boards : Number crunching : GPU version: kernel size tuning and less UI lags (Message 1314)
Posted 2 Dec 2019 by Sergei Chernykh
Post:
Yes, I'll adjust credit for these tasks.
22) Message boards : News : The search up to 10^21 (Message 1308)
Posted 16 Nov 2019 by Sergei Chernykh
Post:
It should work on 1GB GPUs, but I haven't tested it. You'll probably need to reduce kernel size to the minimum.
23) Message boards : Number crunching : GPU version: kernel size tuning and less UI lags (Message 1305)
Posted 14 Nov 2019 by Sergei Chernykh
Post:
Probably because task size is 68*1012, 100 times more than usual. Task generator was definitely not perfect.
24) Message boards : Number crunching : GPU version: kernel size tuning and less UI lags (Message 1303)
Posted 14 Nov 2019 by Sergei Chernykh
Post:
It doesn't get stuck: https://sech.me/boinc/Amicable/workunit.php?wuid=11804634 - another GPU finished it, you just had to wait a bit more.
25) Message boards : Number crunching : GPU version: kernel size tuning and less UI lags (Message 1296)
Posted 10 Nov 2019 by Sergei Chernykh
Post:
It's not GPU lags in your case, you jut need more than 8 GB system memory to run GPU tasks. Your notebook is swapping and everything starts to lag.
26) Message boards : Number crunching : Missing Computers (Message 1289)
Posted 2 Nov 2019 by Sergei Chernykh
Post:
If they weren't active for more than 90 days then they were deleted automatically.
27) Message boards : News : The search up to 10^20 (Message 1286)
Posted 1 Nov 2019 by Sergei Chernykh
Post:
Yes, it's complete. All previously known amicable numbers were found by the search (which was expected), and many new amicable numbers were found.
28) Message boards : News : The search up to 10^21 (Message 1281)
Posted 30 Oct 2019 by Sergei Chernykh
Post:
No, old tasks will just run out soon. You might abort tasks that use v3.01 GPU application, but it's safe to just leave them running without pausing/resuming.
29) Message boards : News : The search up to 10^21 (Message 1278)
Posted 30 Oct 2019 by Sergei Chernykh
Post:
Attention!
There was a bug in GPU version for 1021 search that caused invalid results after resuming from checkpoint in some cases, so I had to fix it and restart the search from the beginning.
30) Message boards : Bug tracker : Constant Computation errors (Message 1276)
Posted 29 Oct 2019 by Sergei Chernykh
Post:
Error -6 is CL_OUT_OF_HOST_MEMORY. Increase swap file size. You need 8.5 GB per each running GPU task.
31) Message boards : Number crunching : Problem with valid wus count on All Tasks page? (Message 1274)
Posted 29 Oct 2019 by Sergei Chernykh
Post:
Yes, there are usually many amicable numbers found per task now because 1021 search started with the most fruitful ranges.
32) Message boards : Number crunching : Prime Numbers (Message 1271)
Posted 28 Oct 2019 by Sergei Chernykh
Post:
If it takes 1 hour to compute prime table, you're doing something seriously wrong. It usually takes no more than 1 minute and is done on a single CPU core, then it starts using all CPU cores, so it's easy to notice how much time it takes.
33) Message boards : News : The search up to 10^20 (Message 1265)
Posted 27 Oct 2019 by Sergei Chernykh
Post:
I think it's different tasks every time. Some 1020 tasks time out after 3 days, new tasks are created, sent and it repeats.

P.S. I was wrong, these tasks were stuck.
34) Message boards : Number crunching : Prime Numbers (Message 1263)
Posted 26 Oct 2019 by Sergei Chernykh
Post:
I prefer not to store prime table on disc as it can get corrupt/overwritten either occasionally or maliciously.
35) Message boards : News : The search up to 10^20 (Message 1250)
Posted 25 Oct 2019 by Sergei Chernykh
Post:
I repeat: these 8 GB can mostly go into swap file without hurting performance. You don't need physical 8 GB RAM per task, OS will swap out unused parts of memory if you have not enough RAM.
36) Message boards : News : The search up to 10^21 (Message 1246)
Posted 25 Oct 2019 by Sergei Chernykh
Post:
The search up to 1021 has started! Over 50,000 new amicable pairs have been found in the first few days.

Due to the huge size of search space, first stage of the search will only look for all amicable pairs where smaller number is of the form 3N*...*p where N > 0 and p < 1011. I expect that this will find 2-3 million new amicable pairs in a year or so.

Note that system memory requirements are higher now because all prime numbers < 1011 need to be stored in memory for the search.
37) Message boards : News : The search up to 10^20 (Message 1245)
Posted 25 Oct 2019 by Sergei Chernykh
Post:
It doesn't require 8 GB all the time, you could increase your swap file and be fine with ~1.5 GB physical RAM per task. Now I'm not sure how to make BOINC run with more memory used than available.
38) Message boards : News : The search up to 10^20 (Message 1235)
Posted 23 Oct 2019 by Sergei Chernykh
Post:
Running GPU on 1021 is fine now. Even 1 task per GPU will get you almost 100% GPU load (except for when task starts), 2 tasks per GPU will guarantee you 100% load all the time. CPU usage of GPU tasks is low, so you can run CPU tasks on the same machine.
39) Message boards : News : The search up to 10^20 (Message 1233)
Posted 21 Oct 2019 by Sergei Chernykh
Post:
When can we expect more of the 10^21 tasks?

As soon as 10^20 tasks run out or a bit before that for a smooth transition.
40) Message boards : News : The search up to 10^20 (Message 1230)
Posted 21 Oct 2019 by Sergei Chernykh
Post:
I pushed search limits for 1021 a bit further into large primes territory, so memory requirements are higher now (all primes up to 1011 are checked). CPU version requires 2.85 GB of system memory. GPU version fits in 1 GB on GPU, but requires 8 GB of system memory after refactoring.
41) Message boards : News : The search up to 10^20 (Message 1227)
Posted 19 Oct 2019 by Sergei Chernykh
Post:
There's literally only 310 such tasks for the entire search. lpp stands for "largest prime power" and when it's 2 or 3, you only get to check sqrt(N) or cuberoot(N) numbers. So we shouldn't bother with them at all. But they are also good for testing and finding bugs, so I added them for the beta test.
42) Message boards : News : The search up to 10^20 (Message 1224)
Posted 19 Oct 2019 by Sergei Chernykh
Post:
I've checked it. Tasks with "/lpp" in command line iterate over many more ranges than normal tasks and many of these ranges end up with 0 numbers for GPU. This is why it uses only CPU in the end and is at 100% - it counts only how many numbers to check on GPU.
43) Message boards : News : The search up to 10^20 (Message 1221)
Posted 19 Oct 2019 by Sergei Chernykh
Post:
Probably task_size parameter was calculated incorrectly, so it can count to 100% before the actual work is done. I'll check this task.
44) Message boards : News : The search up to 10^20 (Message 1219)
Posted 19 Oct 2019 by Sergei Chernykh
Post:
And there is a bug in OpenCL version: https://sech.me/boinc/Amicable/workunit.php?wuid=11571043
I reproduced it locally. Kernel size = 23 skips some numbers, kernel size = 21 works fine. But other work units might behave differently, i.e. kernel size 21 will skip numbers.

P.S. Fixed in version 3.01. It was a bug introduced during refactoring, so 1020 search wasn't affected.
45) Message boards : News : The search up to 10^20 (Message 1218)
Posted 18 Oct 2019 by Sergei Chernykh
Post:
The work units with "/lpp" in command line are special ones, they're more hard on CPU, so 100% GPU load is not guaranteed. But there is only a few of them.

And yes, new pairs: https://sech.me/ap/log/2019/2019-10-19/VOLUNTEERS%40BOINC.txt
46) Message boards : News : The search up to 10^20 (Message 1216)
Posted 18 Oct 2019 by Sergei Chernykh
Post:
If there is some task that doesn't finish on GPU, give me a link to it and I'll check it. And I see that there are already some new pairs found.
47) Message boards : News : The search up to 10^20 (Message 1211)
Posted 18 Oct 2019 by Sergei Chernykh
Post:
I've added applications for 1021, set it as a beta for now and added 1000 first work units for testing. We'll see how it goes.

P.S. But all amicable pairs found during this test will be counted as usual.
48) Message boards : News : The search up to 10^20 (Message 1209)
Posted 17 Oct 2019 by Sergei Chernykh
Post:
Right now it takes 5-10 seconds to generate a task and it's a sequential algorithm - every next range must start where the previous range ended, and I have no way to determine range borders until I run a quick emulation of the search. I can split it into several parallel processes however - for example, N processes for ranges starting with 3N, but each process is also taking 8.5 GB of memory for storing prime numbers. I'll see if I can share these 8.5 GB between processes and run task generators on all 16 threads.

P.S. I think that even now I'm able to generate tasks faster than the crunchers will be able to consume them, so it's not really a problem.
49) Message boards : News : The search up to 10^20 (Message 1207)
Posted 16 Oct 2019 by Sergei Chernykh
Post:
Yes, it was intentional. I need a bit more time to generate work units for 1021. Everything is ready, but task generation is sloooow.... I hope I'll be able to provide around 10,000 WUs next week when the new search starts.
50) Message boards : News : The search up to 10^20 (Message 1204)
Posted 14 Oct 2019 by Sergei Chernykh
Post:
Yes, I'll prepare CPU and GPU applications for 10^21 this week. GPU version will need a bit of refactoring because as of now it requires more than 8 GB of GPU memory for 10^21. I'll have to split workunits into 8 stages (1 GB each) for GPU.


Next 50


©2020 Sergei Chernykh