Page 5 of 6
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: August 19th, 2026, 6:26 pm
by bobinou
Whilst mining for this i have found the currently longest running non agar submitted to the Registry, with a runtime of 2382. This is almost 100 iterations more than my prev. record holder:
Fifty Trillion
Code: Select all
x = 16, y = 16, rule = B3/S23:T16,16
b3o5b2ob2obo$bob2ob3obobo3b$2obo5b2obob2o$obo3b6o2bob$2b2ob3obo3b3o$4bobobob2o2bob$o6bo2bobob2o$3b3obo2bo2bobo$bo2bobob3o4bo$3b2o2b2ob3obob$3bo3bo3b3obo$3o6b5obo$2bo2b2obo2b2obob$bob2o2bobo3b2ob$obob3o7bob$2bo3bo2b3o3bo!
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: August 22nd, 2026, 10:31 am
by hash0
Gentlepeople and GOL lovers,
we are starting the next 6 week long tournament series for GOL patterns on the
16x16 torus with non-repeating generational count between
600 and 699 (B3/S23 ruleset).
The draw for the first weekly series will be after tomorrow,
Sunday, 23rd August, 22:00 CEST, so create your free account today and start mining patterns if you want to participate tomorrow:
https://lifehashes.net/login
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: August 23rd, 2026, 6:29 am
by hash0
Gentlepeople, Game of Life enthusiasts and general computational sports nerds!
We are pleased to announce the schedule for the rest of 2026 for you to peruse, plan, and scheme: ■
https://lifehashes.net/hashwar-schedule ... erview.php

- 2026_weekly_schedule.png (121.02 KiB) Viewed 427 times
Get your Glyphs in order and fire up your miners! We will be playing weight classes 600 and 700 for the remainder of the year. Both classes will be featured in two cycles of three weekly series each, at the end of which we'll be featuring Playoffs to determine the best of the best of the best!
If you want to participate, register yourself a (free!) account at ■
https://lifehashes.net/login and start mining immediately. Here is an explanatory video to walk you through the intricacies of the GUI: ■
https://youtu.be/rFsUK_Bs_V0
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: August 23rd, 2026, 4:24 pm
by hash0
The first Weekly Series for weight class 600 has begun - here is the official drawing of 32 contestant Glyphs (fairness guaranteed by seeding the Fisher-Yates shuffle algorithm with a compressed Randomness Beacon Pulse issued by the U.S. National Institute of Standards and Technology):

- 000 SUNDAY DRAW.png (131.4 KiB) Viewed 422 times
Glyph owners participating this week: Phnx (4), p1geon (4), bobinou (4), Nostromo Eigenmother (3), Fleur-de-Lis (3), YSL (3), R2D2 (2), l0r1ns3r (2), hash0 (1), ObiWanK (1), Vort (1), unname4798 (1), SecondTypist (1), randomaccxd (1), Prometheus (1)
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: August 25th, 2026, 7:29 am
by hash0

- 001 GROUP A RESULTS.png (9.91 KiB) Viewed 406 times
Congratulations to Phnx for fielding the top two Glyphs in this group! Those two and Bobinou's "Cold Ice Tea" already qualified for the weekly finale on Saturday!
The four Glyphs indicated in yellow will have to compete against each other on Friday ("Redemption Day") to dermine who will proceed to the finale. Good luck to SecondTypist, unname4798, p1geon, and hash0 - only one can win!
"Burgundy Bolt", owned by Nostromo Eigenmother, dropped out of the series.
We will be broadcasting Group B live tonight on our twitch channel: ■
https://www.twitch.tv/lifehashes
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: August 30th, 2026, 1:46 pm
by hash0
Line-Up for Series 34 confirmed:

- 000 GROUP DRAWINGS.png (146.99 KiB) Viewed 386 times
Who do you think will win this week?
See the live-streamed drawing (including fetching of the latest Randomness Beacon Pulse issued by the U.S. National Institute for Standards and Technology to ensure fairness) here: ■
https://youtube.com/live/r66mycLL8Xg?feature=share
Follow along the current weekly series: ■
https://lifehashes.net/hashwar-schedule ... ries_id=34
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 5th, 2026, 2:30 am
by hash0
Bracket Draw of the 16 contestants for tonight's weekly Grand Finale:

- 001 FINAL DRAWINGS.png (81.89 KiB) Viewed 366 times
Congratulations specifically to unname4798 ("BLINKER HAVEN") and SecondTypist ("NAME KEPT PG-13") which scored 44,559 and 53,697 points in the Group Phase, respectively.
Follow along the weekly stats here: ■
https://lifehashes.net/hashwar-schedule ... ries_id=34
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 5th, 2026, 10:09 am
by hash0
TONIGHT: Grand Finale of HASHWAR Week 34 *** LIVE ***

- SERIES 34.png (546.8 KiB) Viewed 357 times
Tonight's weekly finale will be cast live on our dedicated channel ■
https://www.youtube.com/@lifehashes
*** 23:00h CEST *** 21:00h UTC *** 5:00 PM EDT *** 4:00 PM CDT *** 2:00 PM PDT ***
Get the current stats for Glyph and Owner records of the perennial series here: ■
https://lifehashes.net/hashwar-schedule ... series.php
All group and redemption matches including the final draw leading of to this weeks finale here: ■
https://lifehashes.net/hashwar-schedule ... ries_id=34
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 18th, 2026, 4:39 am
by hash0

- major_update.png (452.64 KiB) Viewed 314 times
Hashwar Series 700, beginning on 20th September, will be played on the updated Adversarial Conway algorithm, i.e. version 2.0. The updated algorithm contains the following adjustments to the core game mechanic:
■ SELECTION OF GLYPHS FOR 'HASHWAR'
Users must now explicitly flag their Glyphs to be accessible for the weekly group drawings. In the [ GLYPHS_REGISTERED ] tab, all Glyphs have a drop-down menu to choose from. "HW: OFF" (default) means that Glyphs will not be considered for group drawings.
■ CELL DECAY
Cells that were charged within the first 99 generations receive a "charge decay" with increasing probability depending on the longevity of the owning Glyph. This all but levels out scoring differences between Glyphs of the same weight class that have widely differing generational counts. (Consequently, mining for Glyphs with the maximum allowable generational count within the weight class is largely futile.)
■ NEW GAME MECHANIC: 'INTRUSION'
The "Peak" metric determines the number of "shield points" a Glyph has. Shield points are deducted for every overlapping cell when two Glyphs meet on the torus and the opponent Glyph is in OFFENSIVE mode. Once all shield points are deducted, the shieldless Glyph's evolution is halted.
■ NEW PLAYER MECHANIC: 'TACTICAL STANCE'
When making Glyphs accessible for group drawings, players assign their Glyphs one of three "tactical stances" (a "combat mode", so to speak): BALANCED, OFFENSIVE, and DEFENSIVE. Those stances have distinct advantages and disadvantages as follows:
- OFFENSIVE: Will inflict shield damage / trigger 'INTRUSION' when overlapping with the opponent on the torus
- BALANCED: 30% bonus for attacking the opponent's cells
- DEFENSIVE: 100% bonus for charging neutral cells but cannot flip cells belonging to the opponent, higher resistance to Intrusion
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 26th, 2026, 12:33 pm
by hash0

- bobinou_rag.png (498.07 KiB) Viewed 277 times
Bobinou's mining/hashing rig to find optimally dense Conway Glyphs for Hashwar / Gridwar. This is what SERIOUS compute looks like!
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 26th, 2026, 12:38 pm
by NNlk05
hash0 wrote: September 26th, 2026, 12:33 pm
bobinou_rag.png
Bobinou's mining/hashing rig to find optimally dense Conway Glyphs for Hashwar / Gridwar. This is what SERIOUS compute looks like!
Well... since then, I have now gotten access to a 5070...
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 26th, 2026, 12:58 pm
by bobinou
NNlk05 wrote: September 26th, 2026, 12:38 pm
Well... since then, I have now gotten access to a 5070...
That is dope! How well is it performing compared to your prev setup?
The rack is serving for all my compute and storage needs, not just for GOL

and far from finished lol.
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 26th, 2026, 12:59 pm
by LuveelVoom
bobinou wrote: September 26th, 2026, 12:58 pm
The rack is serving for all my compute and storage needs, not just for GOL

and far from finished lol.
We here at conwaylife forums would love to know the specs of this

Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 26th, 2026, 1:13 pm
by bobinou
LuveelVoom wrote: September 26th, 2026, 12:59 pm
We here at conwaylife forums would love to know the specs of this
Well ofc, here are the specs of the current config, it will be filled with more boards and pcs as time goes on.
- Intel nuc N100: visible at the top, is used to host websites, manage the other devices and connected to the display in front
- Jetson Orin Nano Super: for mining boards, ML experiments and hosting mc server from time to time haha
- 2 FPGA boards for mining boards for hash0's challanges
- Luckfox pico ultra: to act as a reciver of the boards which are uploaded and evaluated on a database at glyphs.zip (self advertisement hehe). Mining rate and mode are visible on the 2. screen at the side.
- 4 Bay hotswap NAS with a PI5
- 2 more PI5's for random stuff
The rack itself is mostly 3d printed and designed by myself because it had to be deeper than standard 10" racks.
Things that will be added in the future:
- FPGA cluster
- ITX mini pc
- PI's for home automation
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 26th, 2026, 1:35 pm
by hash0
NNlk05 wrote: September 26th, 2026, 12:38 pm
Well... since then, I have now gotten access to a 5070...
Noice! Wanna put it to good use? I think Vort has written a good script that utilizes GPU parallelization - with that hardware you'd be more than competitive in Hashwar. Give it a try! We are currently playing weight class 700 (16x16 torus, B3/S23, patterns need to live for 700-799 non-repeating generations)! Someone needs to break Bobinou's and Phnx's dominance in HASHWAR!
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 26th, 2026, 1:39 pm
by NNlk05
LuveelVoom wrote: September 26th, 2026, 12:59 pm
bobinou wrote: September 26th, 2026, 12:58 pm
The rack is serving for all my compute and storage needs, not just for GOL

and far from finished lol.
We here at conwaylife forums would love to know the specs of this
NNlk05 wrote: August 19th, 2026, 11:54 pm
Flex
Screenshot_2026-08-19_20-53-40.png
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 26th, 2026, 1:40 pm
by bobinou
hash0 wrote: September 26th, 2026, 1:35 pm
Someone needs to break Bobinou's and Phnx's dominance in HASHWAR!
pff the new rule set has us sweating as well. its hard to find extraordinary boards!
which is a good thing...

Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 26th, 2026, 1:43 pm
by hash0
bobinou wrote: September 26th, 2026, 1:40 pm
hash0 wrote: September 26th, 2026, 1:35 pm
Someone needs to break Bobinou's and Phnx's dominance in HASHWAR!
pff the new rule set has us sweating as well. its hard to find extraordinary boards!
which is a good thing...
Hehe... that's what you get when you apply a little bit of Game Theory to spice things up.
Although I have to say I am seriously impressed by Phnx immediately applying Nash Equilibria analysis to the tactical stance matrix...

Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 27th, 2026, 10:40 pm
by SecondTypist
With my absolutely incarnate skill, I have managed to pull off a “get rekt” moment. I absolutely genuinely with no cheating at all managed to snag 5 of the top 6 results for longest lasting glyph. I definitely did not translate the #1 pattern by a cell at all (yep, totally out of the question) or rotate/reflect it (Yep nothing to see here) or add an inconsequential cell (who would do that?) or construct a grammatically incorrect sentence when boasting about (I’m very skilled at writing you see, and I have never done any that exhibits hubris). Anyway, below is a screenshot of my genuinely legally obtained almost records. (The previous was just a roundabout way of saying that you should probably update how you check for if something is a duplicate.)
Also, on a constructive note, aren’t there programs that can rewind configurations? I think that that seriously hinders your idea of listing a discoverer, as something like that is incentivized, rather than searching for patterns.
Additionally, I would like to comment that no one is ever going to beat me for the glyph that lasts the least generations. You just can’t do it.
3dditionally, can you get rid of your rancid 5-million-posts-in-a-row-on-the-same-exact-thread streaks

. You could edit your previous posts once you start a new one.
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 29th, 2026, 4:37 am
by bobinou
Lol quickly fill it up before its patched xd
I made a thingy to detect duplicate boards even if rotated or shifted and be more than happy if hash0 implements that on his registry.
This would also help the battles to avoid duplicates of the winners competing against each other.
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 29th, 2026, 4:47 am
by Phnx
SecondTypist wrote: September 27th, 2026, 10:40 pm
With my absolutely incarnate skill, I have managed to pull off a “get rekt” moment. I absolutely genuinely with no cheating at all managed to snag 5 of the top 6 results for longest lasting glyph. I definitely did not translate the #1 pattern by a cell at all (yep, totally out of the question) or rotate/reflect it (Yep nothing to see here) or add an inconsequential cell (who would do that?) or construct a grammatically incorrect sentence when boasting about (I’m very skilled at writing you see, and I have never done any that exhibits hubris). Anyway, below is a screenshot of my genuinely legally obtained almost records. (The previous was just a roundabout way of saying that you should probably update how you check for if something is a duplicate.)
Also, on a constructive note, aren’t there programs that can rewind configurations? I think that that seriously hinders your idea of listing a discoverer, as something like that is incentivized, rather than searching for patterns.
Additionally, I would like to comment that no one is ever going to beat me for the glyph that lasts the least generations. You just can’t do it.
3dditionally, can you get rid of your rancid 5-million-posts-in-a-row-on-the-same-exact-thread streaks

. You could edit your previous posts once you start a new one.
Hahahahahaha but still not #1
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 30th, 2026, 1:15 pm
by Phnx
I guess I won
the_creation_of_adam
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 30th, 2026, 3:44 pm
by Phnx
Hello there,
I made something somewhat interesting
https://youtu.be/tOiDHcOxTQE
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 30th, 2026, 3:50 pm
by hash0
I think someone with domain knowledge should take a look at this. Seriously. I am lacking the expertise to accurately gauge the importance of those constructs for the CA/GOL research community - but I am fairly confident that at the very least the classification of periodicals on the 16x16 torus at 5:05 in the video is genuinely interesting (it provides context that there are, indeed, other oscillators beside the p314 Agar that gives rise to the 5,024 generation Glyphs).
Re: "Hashwar": A Battle Royale for individual GOL configurations
Posted: September 30th, 2026, 4:41 pm
by NNlk05
This may look like any apgsearch soup:
Code: Select all
x = 16, y = 16, rule = B3/S23
obobobo3bobo$o4b2ob2o2b2o$4b2ob2o2b2ob2o$2o4b2obo5bo$4o2bobo4bo$ob2o2b
ob2obobo$bo3b2o2b4o$b2obobob2obob2o$o2b2obo2bobo3bo$2ob4ob3ob4o$2bo2b
7ob2o$obo4b2o2bo3bo$2o2b5o2b4o$b3o8bobo$b2ob2obob2ob4o$2b5o2bob5o!
But it isn't:
Code: Select all
x = 16, y = 16, rule = B3/S23:T16,16
obobobo3bobo$o4b2ob2o2b2o$4b2ob2o2b2ob2o$2o4b2obo5bo$4o2bobo4bo$ob2o2b
ob2obobo$bo3b2o2b4o$b2obobob2obob2o$o2b2obo2bobo3bo$2ob4ob3ob4o$2bo2b
7ob2o$obo4b2o2bo3bo$2o2b5o2b4o$b3o8bobo$b2ob2obob2ob4o$2b5o2bob5o!
A taste of what to come, found in about 4 seconds.