mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Software (https://www.mersenneforum.org/forumdisplay.php?f=10)
-   -   mtsieve (https://www.mersenneforum.org/showthread.php?t=23042)

pepi37 2022-07-31 21:28

e:\MTSIEVE\[COLOR=Magenta][B]MTSIEVE-2-3-3[/B][/COLOR]>kbbsieve -P 100000 -W1 -k 77 -b 2 -B 200000
kbbsieve v1.1, a program to find factors of k*b^b+c numbers for fixed k and c = +1/-1
199998 even terms removed
Sieve started: 3 < p < 1e5 with 199998 terms (k = 77, 2 <= b <= 200000) (expecting 180913 factors)
CTRL-C accepted. Threads will stop after sieving to 100000
Decreasing worksize to 5024 since each chunk needs more than 5 seconds to test
Sieve completed at p=100003.
CPU time: 10.28 sec. (0.00 sieving) (1.11 cores)
[U][B]Fatal Error: Something is wrong. Counted terms (22861) != expected terms (22859)[/B][/U]

rogue 2022-08-01 03:53

[QUOTE=pepi37;610574]e:\MTSIEVE\[COLOR=Magenta][B]MTSIEVE-2-3-3[/B][/COLOR]>kbbsieve -P 100000 -W1 -k 77 -b 2 -B 200000
kbbsieve v1.1, a program to find factors of k*b^b+c numbers for fixed k and c = +1/-1
199998 even terms removed
Sieve started: 3 < p < 1e5 with 199998 terms (k = 77, 2 <= b <= 200000) (expecting 180913 factors)
CTRL-C accepted. Threads will stop after sieving to 100000
Decreasing worksize to 5024 since each chunk needs more than 5 seconds to test
Sieve completed at p=100003.
CPU time: 10.28 sec. (0.00 sieving) (1.11 cores)
[U][B]Fatal Error: Something is wrong. Counted terms (22861) != expected terms (22859)[/B][/U][/QUOTE]

I see what the issue is. A factor is counted when it shouldn't be. The output .pfgw file is correct. I will fix sometime this week.

japelprime 2022-08-01 15:41

Maybe this helps pepi37. I was running gfndsievecl. I found that if -g is to low then the CPU will not load any work to GPU. The CPU will do the work without distributing work to GPU. (explains high CPU time and low or no GPU time) when I skip -g command no work goes to GPU because default limit is to low. Not even -g 100 moves the card to workload. When putting -g 200 I start to see the CPU buffering up the extra work to GPU card. From 5% workload for -g 200 to 8% for -g 300. So maybe default -g value should be somehow higher. I used -S 10000

rogue 2022-08-01 18:09

[QUOTE=japelprime;610617]Maybe this helps pepi37. I was running gfndsievecl. I found that if -g is to low then the CPU will not load any work to GPU. The CPU will do the work without distributing work to GPU. (explains high CPU time and low or no GPU time) when I skip -g command no work goes to GPU because default limit is to low. Not even -g 100 moves the card to workload. When putting -g 200 I start to see the CPU buffering up the extra work to GPU card. From 5% workload for -g 200 to 8% for -g 300. So maybe default -g value should be somehow higher. I used -S 10000[/QUOTE]

You could also try a higher value for -G. I haven't tried that myself.

pepi37 2022-08-01 21:10

[QUOTE=japelprime;610617]Maybe this helps pepi37. I was running gfndsievecl. I found that if -g is to low then the CPU will not load any work to GPU. The CPU will do the work without distributing work to GPU. (explains high CPU time and low or no GPU time) when I skip -g command no work goes to GPU because default limit is to low. Not even -g 100 moves the card to workload. When putting -g 200 I start to see the CPU buffering up the extra work to GPU card. From 5% workload for -g 200 to 8% for -g 300. So maybe default -g value should be somehow higher. I used -S 10000[/QUOTE]
I will try this
Thanks!

japelprime 2022-08-14 13:07

I am trying xyyxsievecl. I have some questions.

OpenCL Error: Out of resources
in call to clEnqueueReadBuffer
argument: factorCount

Is this error a driver- old hardware problem or wrong parameter setting ? I am not sure where to look. I did clean driver install.

On command prompt I see "xyyxsieve v1.8" on screen when calling xyyxsievecl version (both versions under same folder)

switch -s b seems not to be working for me here.

When calling xyyxsievecl the program works fine but only for cpu work. -W1 and -W2 are fine but not -G or different -M or any other parameters when I try to effect GPU. Most of the time I get blank screen for 2 second as the card try to start working and then this OpenCL error. When it is stable there is no workrate at GPU when running. I have Geforce GTX 760 card.

The same PC works fine trying gfndsievecl for the gpu part.

rogue 2022-08-15 00:13

This implies you do not have enough GPU memory. I believe there is a -S option. Try reducing the number of steps per GPU execution

japelprime 2022-08-16 20:46

[QUOTE=rogue;611461]This implies you do not have enough GPU memory. I believe there is a -S option. Try reducing the number of steps per GPU execution[/QUOTE]

Thanks. It is working now.

ryanp 2022-08-18 19:11

Another crash. This is with the latest SVN code.

[CODE]$ ./srsieve2 -W 36 -P 1e12 -n 3e6 -N 10e6 -o ferm81_3M_10M.txt -s "81*2^n+1"
srsieve2 v1.6.3, a program to find factors of k*b^n+c numbers for fixed b and variable k and n
(b2) Removed 1750000 algebraic factors for 81*2^n+1 of the form (3^2)*2^(n/2)-3*2^((n+2)/4))+1 when n%4=2
Sieving with generic logic for p >= 3
Sieve started: 3 < p < 1e12 with 5250001 terms (3000000 < n < 10000000, k*2^n+1) (expecting 5041260 factors)
Sieving with single sequence c=1 logic for p >= 257
BASE_MULTIPLE = 30, POWER_RESIDUE_LCM = 720, LIMIT_BASE = 720
Split 1 base 2 sequence into 384 base 2^720 sequences.
Legendre summary: Approximately 2 B needed for Legendre tables
1 total sequences
1 are eligible for Legendre tables
0 are not eligible for Legendre tables
1 have Legendre tables in memory
0 cannot have Legendre tables in memory
0 have Legendre tables loaded from files
1 required building of the Legendre tables
518400 bytes used for congruent q and ladder indices
295200 bytes used for congruent qs and ladders
Segmentation fault[/CODE]

Seems to segfault no matter what value of [C]-P[/C] I try.

rogue 2022-08-18 20:15

[QUOTE=ryanp;611708]Another crash. This is with the latest SVN code.

[CODE]$ ./srsieve2 -W 36 -P 1e12 -n 3e6 -N 10e6 -o ferm81_3M_10M.txt -s "81*2^n+1"
srsieve2 v1.6.3, a program to find factors of k*b^n+c numbers for fixed b and variable k and n
(b2) Removed 1750000 algebraic factors for 81*2^n+1 of the form (3^2)*2^(n/2)-3*2^((n+2)/4))+1 when n%4=2
Sieving with generic logic for p >= 3
Sieve started: 3 < p < 1e12 with 5250001 terms (3000000 < n < 10000000, k*2^n+1) (expecting 5041260 factors)
Sieving with single sequence c=1 logic for p >= 257
BASE_MULTIPLE = 30, POWER_RESIDUE_LCM = 720, LIMIT_BASE = 720
Split 1 base 2 sequence into 384 base 2^720 sequences.
Legendre summary: Approximately 2 B needed for Legendre tables
1 total sequences
1 are eligible for Legendre tables
0 are not eligible for Legendre tables
1 have Legendre tables in memory
0 cannot have Legendre tables in memory
0 have Legendre tables loaded from files
1 required building of the Legendre tables
518400 bytes used for congruent q and ladder indices
295200 bytes used for congruent qs and ladders
Segmentation fault[/CODE]

Seems to segfault no matter what value of [C]-P[/C] I try.[/QUOTE]

Okay. I will look into it. Does this happen with lower values for -W? I suspect that switch is behind this.

ryanp 2022-08-18 20:57

[QUOTE=rogue;611710]Okay. I will look into it. Does this happen with lower values for -W? I suspect that switch is behind this.[/QUOTE]

Yeah, with [C]-W 4[/C] it seems to be working.


All times are UTC. The time now is 16:57.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.