Luveel's compute vault

A forum for topics that don't fit elsewhere. Introduce yourselves to other members of the forums, discuss how your name evolves when written out in the Game of Life, or just tell us how you found it. Forum rules still apply.
Post Reply
User avatar
LuveelVoom
Posts: 532
Joined: April 27th, 2022, 7:59 pm

Luveel's compute vault

Post by LuveelVoom »

What is this?

Usually, I'm running some sort of search overnight. However, when I'm going to go to sleep and I don't have anything to search, my computer is just idling there. Imagine all the wasted cycles! So if I don't have something to do with them, I need to give them to you guys.

My computer has 10 performance cores and 48gb ram (although about a fourth of that is used by system processes, so assume 36gb). I can run the following programs:

-qfind (Haven't tested it yet but it installed correctly) Edit: I did some reading and it seems like it might be tricky to figure out how to multithread this. I'll try and do it, but any search with it will be very slow at the moment)
-apgsearch
-EPE
-LLSSS, as well as gchromodynamic's C++ port, which is faster but does not support floating rows and some other features
-VLS (speedydelete's version of LLS, which is currently slower but will likely become faster whenever she implements CDCL; it also can accept barrister patterns as inputs). I might try and install LLS at some point too
-Some other ones which I don't remember because I'm not at my PC right now
-I'll try and install anything

What I cannot do:

-ikpx2 (it doesn't compile correctly on my computer)
-One other one that I forgot (I'll update this post when I'm at my PC)

How do I get this compute time?

Reply to this thread with the description of what you want computed. If I find it interesting, I'll put it in a queue to use whenever I don't have anything to run overnight. Since I have 10 cores, you'll get about 60-80 CPU hours. I'll like your post if I found it interesting enough to put in the queue. If it's particularly interesting, I might keep running it during the day. The queue will probably be updated in this post, but I might put it in a doc instead depending on how I'm feeling later today

More details/effort/explanation of why it's important put into the post = more likely!

Have fun :D
User avatar
TheWayOfTheCon
Posts: 280
Joined: March 28th, 2025, 11:40 pm
Location: Kraken Mare, Titan

Re: Luveel's compute vault

Post by TheWayOfTheCon »

I would like D4_x4 and D8_1 hauls of B34/S34-ae.

The reason is that I have found this rule kinda interesting, as, unlike 3-4 Life, it is not explosive and actually somewhat chaotic. It also has a common c/5d, about as common as Life's glider.

I think with the right high-period oscillator, a gun for the spaceship could be constructed.
I could've chose a better username, but oh well.

Still learning the ropes of cellular automata, focused on one OCA at a time. My current interest is B35/S126 and range-two LTLs.
User avatar
NNlk05
Posts: 614
Joined: January 14th, 2026, 8:42 pm
Location: Exploring in the Jungle of the INT Rulespace
Contact:

Re: Luveel's compute vault

Post by NNlk05 »

D2_1 Blogic. A very interesting rule that has great potential for engineering.
Last edited by NNlk05 on September 7th, 2026, 1:55 pm, edited 1 time in total.
Feci quod potui, faciant meliora potentes.

Code: Select all

x = 10, y = 3, rule = B34twz/S23
b2o4b2o$obo4bobo$2bo4bo!
[[ AUTOSTART AUTOHIDEGUI TRACK 0 -47/270 ZOOM 4 GPS 45 STEP 3 THEME BOOK ]]
https://nnlk05.github.io

=3
User avatar
LuveelVoom
Posts: 532
Joined: April 27th, 2022, 7:59 pm

Re: Luveel's compute vault

Post by LuveelVoom »

TheWayOfTheCon wrote: September 7th, 2026, 1:40 pm I would like D4_x4 and D8_1 hauls of B34/S34-ae.

The reason is that I have found this rule kinda interesting, as, unlike 3-4 Life, it is not explosive and actually somewhat chaotic. It also has a common c/5d, about as common as Life's glider.

I think with the right high-period oscillator, a gun for the spaceship could be constructed.
Are you the anonymous user who has been apgsearching this rule? https://catagolue.hatsya.com/census/b34s34-ae?
They have found an xp40, xp12, xp14, xp18, and xq29:

Code: Select all

x = 13, y = 69, rule = B34/S34-ae
3bo2bo$2b2o2b2o$2b2o2b2o$3bo2bo11$b3o5b3o$2bo7bo$3b2o3b2o$2b3o3b3o$2b
obo3bobo$2b3o3b3o$3b2o3b2o$2bo7bo$b3o5b3o8$4b4obo$4bob2o$2bo4bo$4bob2o
$4b4obo9$5b3o$5b3o$5b3o2$6bo$3o7b3o$3obo3bob3o$3o7b3o$6bo2$5b3o$5b3o$
5b3o9$7bo$5b4o$5bo2bo$5b4o$7bo!
Last edited by LuveelVoom on September 7th, 2026, 1:57 pm, edited 1 time in total.
User avatar
LuveelVoom
Posts: 532
Joined: April 27th, 2022, 7:59 pm

Re: Luveel's compute vault

Post by LuveelVoom »

NNlk05 wrote: September 7th, 2026, 1:51 pmD2_1 Blogic
The reason I've been holding off on searching blogic is because the rulespace is still being explored and the original rule doesn't seem that interesting. I'll do some apgsearching once a more interesting rule in the rulespace is decided on as a replacement for the original.
User avatar
TheWayOfTheCon
Posts: 280
Joined: March 28th, 2025, 11:40 pm
Location: Kraken Mare, Titan

Re: Luveel's compute vault

Post by TheWayOfTheCon »

LuveelVoom wrote: September 7th, 2026, 1:54 pm Are you the anonymous user who has been apgsearching this rule? https://catagolue.hatsya.com/census/b34s34-ae?
They have found an xp40 and xq29:

Code: Select all

x = 13, y = 26, rule = B34/S34-ae
5b3o$5b3o$5b3o2$6bo$3o7b3o$3obo3bob3o$3o7b3o$6bo2$5b3o$5b3o$5b3o9$7bo
$5b4o$5bo2bo$5b4o$7bo!
Yeah, I contribute anonymously on Catagolue.

The period-29 spaceship is somewhat common, actually. And the period-40 oscillator is just four of them bumping into each other.
I could've chose a better username, but oh well.

Still learning the ropes of cellular automata, focused on one OCA at a time. My current interest is B35/S126 and range-two LTLs.
User avatar
J5L
Posts: 51
Joined: July 31st, 2026, 8:53 am
Location: Flower Garden

Re: Luveel's compute vault

Post by J5L »

Please do both B3-kq4j5k6-k/S2-i34y5ai6k and B3-kq4j5k6-k/S2-i35ai6k on C1, D2_1, D4_1, and D8_1.
EDIT 1: removed B3kq since they are already explosive with B3kq
Last edited by J5L on September 9th, 2026, 6:09 am, edited 1 time in total.
JayFiveEll.
Recovering from LWTDS.
User avatar
MDA
Posts: 236
Joined: June 8th, 2026, 12:18 pm
Location: In Haven (B3-ry4acenqt5eir6-ek/S2-a3-a4nq5aeknr6-ak)
Contact:

Re: Luveel's compute vault

Post by MDA »

J5L wrote: September 9th, 2026, 4:17 am Please do both B34j5k6-k/S2-i34y5ai6k and B34j5k6-k/S2-i35ai6k [...]
Both of those rules are explosive and cannot be apgsearched.
My website (a database of Life objects, WIP).
GlidINT has been released!

Code: Select all

x = 5, y = 17, rule = B3-ry4acenqt5eir6-ek/S2-a3-a4nq5aeknr6-akHistory
C$2C$.2C$2C11$.3D$3.D$2.3D!
[[ STOP 72 ]]
User avatar
J5L
Posts: 51
Joined: July 31st, 2026, 8:53 am
Location: Flower Garden

Re: Luveel's compute vault

Post by J5L »

MDA wrote: September 9th, 2026, 5:25 am
J5L wrote: September 9th, 2026, 4:17 am Please do both B34j5k6-k/S2-i34y5ai6k and B34j5k6-k/S2-i35ai6k [...]
Both of those rules are explosive and cannot be apgsearched.
Whoops, forgot to remove B3kq.
JayFiveEll.
Recovering from LWTDS.
Disaster16439
Posts: 331
Joined: June 30th, 2023, 9:17 am
Location: Teyvat

Re: Luveel's compute vault

Post by Disaster16439 »

c/6o searches in B34q/S23-k in LLSSS/qfind (any is fine, I'm not completely sure which is better. In terms of symmetry, it seems like C1 and D2_2 are the best). Note that I've done ikpx2 up to w8, although I'm not sure because different programs have different widths...

w10/11 would be nice. Just for reference if you're curious for other speeds, there are c/2, c/3, c/4, c/5, 2c/5, c/10, c/4d and c/5d already, with c/5o and 2c/5o being ones that could be reduced (with 2c/5 needing a 30P5H2V0-style approach, asymetriccal).

Edit:

I meant w10. w9 I'm pretty sure I've zfinded. Also currently grinding away at some random 3c/8 search. Ram usage from zfind alone: 5.75GB/8GB. Yeah, this is why I wanted this to get qfinded.

Code: Select all

x=0,y=0,rule=B34q/S23-k
14b3o$13bo3bo$13b2ob2o9$15bo$15bo$b2o12bo12b2o$obo25bobo$o10b3o3b3o10b
o$obo25bobo$b2o12bo12b2o$15bo$15bo9$13b2ob2o$13bo3bo$14b3o!
[[ LOOP 200 THEME POISON AUTOSTART T 0 PAUSE 0.3 ]]
I’m sandless :D
User avatar
LuveelVoom
Posts: 532
Joined: April 27th, 2022, 7:59 pm

Re: Luveel's compute vault

Post by LuveelVoom »

Will look into the rules proposed shortly; I may have open time tonight.
User avatar
LuveelVoom
Posts: 532
Joined: April 27th, 2022, 7:59 pm

Re: Luveel's compute vault

Post by LuveelVoom »

Current queue for whenever I get an open night:
D4_x4 and D8_1 hauls of B34/S34-ae (Will likely receive 2 threads each) Done! Will leave running while I'm away from home today.
D2_1 hauls of Blogic (Will likely receive 3 threads) Done! Will leave running while I'm away from home today.
c/6o in B34q/S23-k (Will likely receive 3 threads) No results at w12 even (LLSSS, not qlife). Largest partial:

Code: Select all

x = 18, y = 68, rule = B34q/S23-k
8b2o$6bo4bo$5bo6bo$4bo8bo$4b2ob4ob2o3$6b2o2b2o$7bo2bo$7bo2bo$2b2o2bo4b
o2b2o$2bob2o6b2obo$3bobo6bobo$4bo8bo$2bobo8bobo2$bobo10bobo$bo14bo2$2b
o2b3o2b3o2bo$2bob4o2b4obo$bo3bo6bo3bo$bo3bo6bo3bo$2bobo8bobo$2bo12bo$
3bobo6bobo$3bobo6bobo$2b3o8b3o$bo14bo$o2b4o4b4o2bo$o5bo4bo5bo$6bo4bo$
2bo2bo6bo2bo$5bo6bo$3b2o8b2o$2bo2b8o2bo$6b6o2$5bob4obo$4b2o6b2o$4bo2b
o2bo2bo$5b2o4b2o$5b2o4b2o2$3b12o$b2o2bo2b2o2bo2b2o$bo2bo3b2o3bo2bo2$b
2o3bo4bo3b2o$4b3o4b3o$4bo3b2o3bo$bo4b6o4bo$bob2ob2o2b2ob2obo$7b4o$4bo
2bo2bo2bo$3bo3bo2bo3bo$2bo12bo$2b2o3b4o3b2o$b3o3b4o3b3o$o4b2o4b2o4bo$
o3bob2o2b2obo3bo$bo6b2o6bo$2b3o8b3o$2bo3b2o2b2o3bo$2o3b3o2b3o3b2o$2o3b
3o2b3o3b2o$6bob2obo$3b2o3b2o3b2o!
May try w13, depending on how much memory w12 ended up using.

I'm holding off on the B3-kq rules because I don't want to devote 8 of the 10 threads I'm going to use to one task.
Last edited by LuveelVoom on September 10th, 2026, 11:11 am, edited 1 time in total.
User avatar
LuveelVoom
Posts: 532
Joined: April 27th, 2022, 7:59 pm

Re: Luveel's compute vault

Post by LuveelVoom »

LuveelVoom wrote: September 9th, 2026, 11:13 pm Current queue for whenever I get an open night:
D4_x4 and D8_1 hauls of B34/S34-ae (Will likely receive 2 threads each)
D2_1 hauls of Blogic (Will likely receive 3 threads)
c/6o in B34q/S23-k (Will likely receive 3 threads)

I'm holding off on the B3-kq rules because I don't want to devote 8 of the 10 threads I'm going to use to one task.
Searches have begun!
User avatar
J5L
Posts: 51
Joined: July 31st, 2026, 8:53 am
Location: Flower Garden

Re: Luveel's compute vault

Post by J5L »

J5L wrote: September 9th, 2026, 4:17 am Please do both B3-kq4j5k6-k/S2-i34y5ai6k and B3-kq4j5k6-k/S2-i35ai6k on C1, D2_1, D4_1, and D8_1.
[snip]
Can you do these rules?
JayFiveEll.
Recovering from LWTDS.
User avatar
LuveelVoom
Posts: 532
Joined: April 27th, 2022, 7:59 pm

Re: Luveel's compute vault

Post by LuveelVoom »

Disaster16439 wrote: September 9th, 2026, 8:16 am c/6o searches in B34q/S23-k in LLSSS/qfind (any is fine, I'm not completely sure which is better. In terms of symmetry, it seems like C1 and D2_2 are the best).
Tried a search using llsss-ng w13 D2_2. It got extremely deep. I terminated it for my computer's safety; the program outputs gargantuan save files and I was afraid it would be over 100 gigabytes in size. Here was the last partial:

Code: Select all

x = 20, y = 129, rule = B34q/S23-k
9b2o$7bo4bo$6bo6bo$5bo8bo$5b2ob4ob2o3$7b2o2b2o$7b2o2b2o3$4b2o8b2o$5bo
b6obo$4b3o6b3o$3b3o8b3o$2bo2bobo4bobo2bo$3bo5b2o5bo$7bo4bo$8b4o$6bobo
2bobo$5b2o6b2o$5bob6obo$2bo3bo6bo3bo$bob2o10b2obo$bobo5b2o5bobo$9b2o2$
b2o14b2o$ob2o12b2obo$o3bo10bo3bo$3bo2bo6bo2bo$5b2o6b2o$3b2o2bo4bo2b2o
$2bobo10bobo$6bo6bo$4b2o8b2o$4b3o6b3o$5bo8bo$b2obobo6bobob2o$3b2obo6b
ob2o$bo2bo10bo2bo2$obo14bobo$bo16bo$3b5o4b5o$bobo4bo2bo4bobo$bo6b4o6b
o$4bob2o4b2obo$4bob2ob2ob2obo$4b2o3b2o3b2o$4bo4b2o4bo$3b2o10b2o$3b2o10b
2o$3b5o4b5o$3bobo8bobo$2b2o12b2o$b2o14b2o$6bo6bo$5b2o6b2o$3b2o10b2o$3b
3o8b3o$6b2o4b2o$7bo4bo$5b2o6b2o$5bo8bo$4bo2bo4bo2bo$3bob3o4b3obo$3bo3b
2o2b2o3bo$2bo2bob2o2b2obo2bo$2b5o6b5o2$2b2o12b2o$bo2bo10bo2bo$bobobo8b
obobo$2b2obo2b4o2bob2o$4b2obo4bob2o$5bo8bo$b2o3b2o4b2o3b2o$b2o3b2o4b2o
3b2o$b2o14b2o$2bo14bo$bob5o4b5obo$2bo4bo4bo4bo$bo3bobob2obobo3bo$bo3b
o3b2o3bo3bo$bo3b3o4b3o3bo$5b3o4b3o$3bo12bo$7bo4bo$4bo4b2o4bo$9b2o$5b3o
4b3o$5bo8bo$4bo10bo$3bo12bo$2b2o12b2o$2bobo10bobo$b3ob10ob3o$bo4b8o4b
o$5bo3b2o3bo$4bobo2b2o2bobo$5b3ob2ob3o$4bo4b2o4bo$4bo2bo4bo2bo$3bo2bo
6bo2bo$4bo10bo$4b2o8b2o$3bo12bo$6bo6bo$4b4o4b4o$2bo3bobo2bobo3bo$bo6b
4o6bo$2bo14bo$b2obobo2b2o2bobob2o$3bo12bo$3bo3b2o2b2o3bo$b2o5b4o5b2o$
b6obo2bob6o$2bo5bo2bo5bo$4bo2bo4bo2bo$2bobobo2b2o2bobobo$2bo5b4o5bo$b
o5bo4bo5bo2$4b3o6b3o$bob2o3b4o3b2obo$obobobobo2bobobobobo$2bo2bobo4bo
bo2bo$o6bo4bo6bo!
I'm doing a recentering search now.

Edit: I'm an idiot. The save files get smaller as the search deepens. There goes 160 CPU hours...
Post Reply