apgsearch v5.0

For general discussion about Conway's Game of Life.
GUYTU6J
Posts: 2200
Joined: August 5th, 2016, 10:27 am
Location: 拆哪!I repeat, CHINA! (a.k.a. 种花家)
Contact:

Re: apgsearch v5.0

Post by GUYTU6J » July 11th, 2019, 12:40 pm

Could I retrieve the hauls that failed to upload due to bad internet connection and were terminated with q?
My two hours' electricity is to be wasted:

Code: Select all

Greetings, this is apgluxe v5.09-ll2.2.12, configured for b38s23/C1.
Lifelib version: ll2.2.12
Compiler version: 7.4.0
Python version: '3.6.8 (default, Feb 14 2019, 22:09:48)  [GCC 7.4.0]'
[omitted]
37233589 soups completed.
Attempting to contact payosha256.
Internet does not exist; henceforth trying gol.hatsya.co.uk instead of catagolue.appspot.com...
No response!
Connection was unsuccessful.
"Fortunately", no new discovery, and 37233589 is prime.

User avatar
testitemqlstudop
Posts: 1368
Joined: July 21st, 2016, 11:45 am
Location: in catagolue
Contact:

Re: apgsearch v5.0

Post by testitemqlstudop » July 11th, 2019, 10:43 pm

No, unless you logged your hauls.

With a script you can "re-upload" logged-but-unsubmitted hauls.

GUYTU6J
Posts: 2200
Joined: August 5th, 2016, 10:27 am
Location: 拆哪!I repeat, CHINA! (a.k.a. 种花家)
Contact:

Re: apgsearch v5.0

Post by GUYTU6J » July 11th, 2019, 11:09 pm

testitemqlstudop wrote:No, unless you logged your hauls.With a script you can "re-upload" logged-but-unsubmitted hauls.
How to log and where's the script?

wildmyron
Posts: 1542
Joined: August 9th, 2013, 12:45 am
Location: Western Australia

Re: apgsearch v5.0

Post by wildmyron » July 12th, 2019, 1:36 am

Unfortunately, as testitemqlstudop said, those results are not recoverable. Logging the submitted hauls with apgsearch can be done by giving the option "-L 1" to apgluxe. Normally this is combined with the testing option "-t 1" which disables uploading to Catagolue. However, the way that apgluxe deals with no internet connection is to continue with the current haul using the same seed and adding to the local census. This means it doesn't write a log file when it continues the search after authentication failed (if it did, then successive log files would contain duplicated results). Unfortunately this means that when apgluxe is quit, but authentication fails, no log file is written.

@calcyman:

- Is it possible to modify SoupSearcher.submitResults() to save the log file when apgluxe has been quit? I guess this would mean passing "running" (or "finished") into submitResults()

- Also a bug in option processing: "-t 1 -i 2 -L 1" causes apgluxe to run two iterations, but "-i 2 -t 1 -L 1" only does one iteration. Perhaps you could make the following change:

Code: Select all

--- a/includes/apgluxe.h
+++ b/includes/apgluxe.h
@@ -79,7 +79,7 @@ int run_apgluxe(int argc, char *argv[]) {
             unicount = atoi(argv[i+1]);
         } else if (strcmp(argv[i], "-t") == 0) {
             testing = atoi(argv[i+1]);
-            if (testing) { iterations = 1; }
+            if (testing && !iterations) { iterations = 1; }
         } else if (strcmp(argv[i], "-p") == 0) {
             parallelisation = atoi(argv[i+1]);
         } else if (strcmp(argv[i], "--rule") == 0) {
The 5S project (Smallest Spaceships Supporting Specific Speeds) is now maintained by AforAmpere. The latest collection is hosted on GitHub and contains well over 1,000,000 spaceships.

Semi-active here - recovering from a severe case of LWTDS.

GUYTU6J
Posts: 2200
Joined: August 5th, 2016, 10:27 am
Location: 拆哪!I repeat, CHINA! (a.k.a. 种花家)
Contact:

Re: apgsearch v5.0

Post by GUYTU6J » July 22nd, 2019, 2:20 am

How about displaying current integrated haul size on the apgluxe console together with search speed and extent?
EDIT example: a line would be

Code: Select all

b3s23/C1: xxxx soups completed (xxxx soups/second current, xxxx overall), xxxx objects (xxxx distinct) found (total size xxx KB)
Last edited by GUYTU6J on July 22nd, 2019, 9:47 am, edited 1 time in total.

User avatar
testitemqlstudop
Posts: 1368
Joined: July 21st, 2016, 11:45 am
Location: in catagolue
Contact:

Re: apgsearch v5.0

Post by testitemqlstudop » July 22nd, 2019, 5:13 am

What do you mean

GUYTU6J
Posts: 2200
Joined: August 5th, 2016, 10:27 am
Location: 拆哪!I repeat, CHINA! (a.k.a. 种花家)
Contact:

Re: apgsearch v5.0

Post by GUYTU6J » July 23rd, 2019, 1:10 am

testitemqlstudop wrote:
testitemqlstudop wrote:Bump.
testitemqlstudop wrote:How difficult would it be to modify apgsearch so that it skips soups that can't be censused in 10 seconds? I'm searching a b2a rule that has soup search speed swings (alliteration not intended) from 1200 soups/second to 0.012 soups/second (no kidding.) The culprit seems to be in yl detection, especially for very high yl periods - this b2a rule lacks replicators.

Alternatively, what would happen if I changed the "maximum pathological attempts" from 5 to something like 2?

YL detection is especially problematic, even in rules like B2ei3aij/S23aijny when all there is is a four-barreled gun.
Seconded.Searching my B34twyzS23-ekn4eit is painfully slow and sometimes freezes.
EDIT: it would be better if we can input allowed maximum census time in the command, and see a

Code: Select all

xxx soups discarded overall
message immediately after a soup is skipped.Or displaying the problematic soup pattern directly so that we can copy it to golly and run it manually...
Last edited by GUYTU6J on July 25th, 2019, 10:50 pm, edited 1 time in total.

User avatar
testitemqlstudop
Posts: 1368
Joined: July 21st, 2016, 11:45 am
Location: in catagolue
Contact:

Re: apgsearch v5.0

Post by testitemqlstudop » July 23rd, 2019, 6:09 am

Thank you!

GUYTU6J
Posts: 2200
Joined: August 5th, 2016, 10:27 am
Location: 拆哪!I repeat, CHINA! (a.k.a. 种花家)
Contact:

Re: apgsearch v5.0

Post by GUYTU6J » July 25th, 2019, 10:34 pm

Another request: lifting threshold for megasized patterns in b3s23/C4_4 automatically.
This is annoying:

Code: Select all

b3s23/C4_4: 107735 soups completed (3483.352 soups/second current, 3590.927 overall).
Soup k_EQSUyCnLstx8108612 has a final population of 3532 cells.
Soup k_EQSUyCnLstx8109219 has a final population of 3044 cells.
Soup k_EQSUyCnLstx8110114 has a final population of 3568 cells.
Soup k_EQSUyCnLstx8110808 has a final population of 3740 cells.
Soup k_EQSUyCnLstx8117371 has a final population of 3064 cells.
Soup k_EQSUyCnLstx8117668 has a final population of 3136 cells.
Soup k_EQSUyCnLstx8120478 has a final population of 3280 cells.
Soup k_EQSUyCnLstx8121030 has a final population of 3768 cells.
Soup k_EQSUyCnLstx8123879 has a final population of 4428 cells.
Soup k_EQSUyCnLstx8125381 has a final population of 3060 cells.
Soup k_EQSUyCnLstx8127892 has a final population of 3164 cells.
Soup k_EQSUyCnLstx8132903 has a final population of 3584 cells.
Soup k_EQSUyCnLstx8133617 has a final population of 3500 cells.
Soup k_EQSUyCnLstx8136766 has a final population of 4780 cells.
Soup k_EQSUyCnLstx8142934 has a final population of 3264 cells.
b3s23/C4_4: 143204 soups completed (3546.900 soups/second current, 3579.832 overall).
I guess setting the minimum to 4500 cells is preferable.

GUYTU6J
Posts: 2200
Joined: August 5th, 2016, 10:27 am
Location: 拆哪!I repeat, CHINA! (a.k.a. 种花家)
Contact:

Re: apgsearch v5.0

Post by GUYTU6J » August 4th, 2019, 11:51 am

Help! After some multitasking on my computer, my apgsearch has dramatically slowed down to about 2000 soups/sec with B3S23/C1 even when I only started apgsearch. Its CPU usage decreases too. How to fix this?
How long will I be free from being in suspense(suspesion?) without any replies in this thread?

User avatar
muzik
Posts: 5578
Joined: January 28th, 2016, 2:47 pm
Location: Scotland

Re: apgsearch v5.0

Post by muzik » September 15th, 2019, 7:53 am

Since apgsearch can now support genext rules for range 1, and it can support Generations rule of range >1, can it be made to support genext rules of range >1?

User avatar
Ian07
Moderator
Posts: 891
Joined: September 22nd, 2018, 8:48 am
Location: New Jersey, US

Re: apgsearch v5.0

Post by Ian07 » October 3rd, 2019, 4:33 pm

apgsearch apparently gets confused and marks this pseudo-flotilla as a true object if there's another spaceship close enough behind it:

Code: Select all

x = 16, y = 31, rule = B3/S23
boooobbboobooboo$
oooboobbobooboob$
obbboboobboooobb$
bboooboboobboobb$
boobobbobobooooo$
boobobooobbboboo$
booobbbbboobbobo$
bobbbobobbbbobbb$
bbbooobbooooooob$
obobbbooboobbobo$
obobobobboobobbb$
obbbbbbobobbooob$
bbbobboboboobbob$
obbbooboobbbbobo$
oobobboobbbboobo$
oboboboboobbobob$
oobobboobbbboobo$
obbbooboobbbbobo$
bbbobboboboobbob$
obbbbbbobobbooob$
obobobobboobobbb$
obobbbooboobbobo$
bbbooobbooooooob$
bobbbobobbbbobbb$
booobbbbboobbobo$
boobobooobbboboo$
boobobbobobooooo$
bboooboboobboobb$
obbboboobboooobb$
oooboobbobooboob$
boooobbboobooboo!

Code: Select all

x = 31, y = 31, rule = B3/S23
ooboooboooboooooooooboooboooboo$
booooboobbbobooboobobbbooboooob$
bbooboobooooobobobooooobooboobb$
bobbbobbobbobooooobobbobbobbbob$
oooobooboooobobbboboooobooboooo$
bboobbbobooooobobooooobobbboobb$
bbooobbobbooobobobooobbobbooobb$
bbobbooooobobbobobbobooooobbobb$
oobbobbbboobobbobboboobbbbobboo$
bbooboooooobobbbbbobooooooboobb$
booobbbobbobooooooobobbobbbooob$
ooooobobbbbobbooobbobbbbobooooo$
ooobboboobbbbooboobbbboobobbooo$
obboobbbbobooooboooobobbbboobbo$
obbboooooooobbooobboooooooobbbo$
bboobbboooobobbobboboooobbboobb$
obbboooooooobbooobboooooooobbbo$
obboobbbbobooooboooobobbbboobbo$
ooobboboobbbbooboobbbboobobbooo$
ooooobobbbbobbooobbobbbbobooooo$
booobbbobbobooooooobobbobbbooob$
bbooboooooobobbbbbobooooooboobb$
oobbobbbboobobbobboboobbbbobboo$
bbobbooooobobbobobbobooooobbobb$
bbooobbobbooobobobooobbobbooobb$
bboobbbobooooobobooooobobbboobb$
oooobooboooobobbboboooobooboooo$
bobbbobbobbobooooobobbobbobbbob$
bbooboobooooobobobooooobooboobb$
booooboobbbobooboobobbbooboooob$
ooboooboooboooooooooboooboooboo!

Code: Select all

x = 31, y = 31, rule = B3/S23
bbbbooobbbbbbobbbobbbbbbooobbbb$
bboooooobbbbooobooobbbboooooobb$
booobbbooobobobbbobobooobbbooob$
booobbobobbobbooobbobbobobbooob$
oobbbboboooooobbboooooobobbbboo$
oobbbbbbbooobbobobbooobbbbbbboo$
ooboobboobobbbooobbboboobbooboo$
boobbbobbooboboboboboobbobbboob$
bbooobobobbobobbbobobbobobooobb$
bboboobobbobooobooobobboboobobb$
bbbbooooboobbbbobbbbooboooobbbb$
bboooobbobbobobobobobbobboooobb$
bobbobbobobbbooboobbbobobbobbob$
ooobobbbooboobobobooboobbbobooo$
bobobooobobbooooooobbobooobobob$
bbbobbobbboobbobobboobbbobbobbb$
bobobooobobbooooooobbobooobobob$
ooobobbbooboobobobooboobbbobooo$
bobbobbobobbbooboobbbobobbobbob$
bboooobbobbobobobobobbobboooobb$
bbbbooooboobbbbobbbbooboooobbbb$
bboboobobbobooobooobobboboobobb$
bbooobobobbobobbbobobbobobooobb$
boobbbobbooboboboboboobbobbboob$
ooboobboobobbbooobbboboobbooboo$
oobbbbbbbooobbobobbooobbbbbbboo$
oobbbboboooooobbboooooobobbbboo$
booobbobobbobbooobbobbobobbooob$
booobbbooobobobbbobobooobbbooob$
bboooooobbbbooobooobbbboooooobb$
bbbbooobbbbbbobbbobbbbbbooobbbb!

Code: Select all

x = 31, y = 31, rule = B3/S23
obbbbbbbooooboobooboooobbbbbbbo$
booboobbooobbobbbobbooobbooboob$
bobbobboobbbbooboobbbboobbobbob$
bbboobobbbooobobobooobbboboobbb$
booooboobobbooboboobbobooboooob$
bobbbbbooooobobbbobooooobbbbbob$
bbboobbooobooboooboobooobboobbb$
bbobooooobobobobobobobooooobobb$
ooobbooooobbbbobobbbbooooobbooo$
oobboooboboobbbbbbboobobooobboo$
ooboboboboooobobobooooboboboboo$
obboboobboobbbobobbboobboobobbo$
bbbooboobboboobbboobobbooboobbb$
oooboobbbbbboobbboobbbbbboobooo$
oboobboooboobbbobbboobooobboobo$
bbbbobobbbbbbbobobbbbbbbobobbbb$
oboobboooboobbbobbboobooobboobo$
oooboobbbbbboobbboobbbbbboobooo$
bbbooboobboboobbboobobbooboobbb$
obboboobboobbbobobbboobboobobbo$
ooboboboboooobobobooooboboboboo$
oobboooboboobbbbbbboobobooobboo$
ooobbooooobbbbobobbbbooooobbooo$
bbobooooobobobobobobobooooobobb$
bbboobbooobooboooboobooobboobbb$
bobbbbbooooobobbbobooooobbbbbob$
booooboobobbooboboobbobooboooob$
bbboobobbbooobobobooobbboboobbb$
bobbobboobbbbooboobbbboobbobbob$
booboobbooobbobbbobbooobbooboob$
obbbbbbbooooboobooboooobbbbbbbo!

gama
Posts: 17
Joined: October 24th, 2018, 7:14 am
Location: In the universe

Re: apgsearch v5.0

Post by gama » January 1st, 2020, 11:23 am

bump

I don't know if this is appropriate to post here, but how I get rid of this error?

Code: Select all

Greetings, this is apgluxe v5.09-ll2.2.22, configured for b3s23/C1.

Lifelib version: ll2.2.22
Compiler version: 7.4.0
Python version: '3.6.9 (default, Jul 21 2019, 14:33:59)  [GCC 7.4.0]'

Peer-reviewing hauls:

No more hauls to verify.

Peer-review complete; proceeding search.

Using seed k_nmpmmh9xTkTJ
Instruction set SSE3 detected
Illegal instruction (core dumped)

Code: Select all

x = 25, y = 26, rule = Crawlers
21.A2$19.A2.A.A$22.ABA$17.A2.A$20.ABA$15.A2.A$18.ABA$13.A2.A$16.ABA$
11.A2.A$14.ABA$9.A2.A$12.ABA$7.A2.A$10.ABA$5.A2.A$8.ABA$3.A2.A$6.ABA$
.A2.A$4.ABA$2.A$2.ABA2$2BA!

User avatar
Hdjensofjfnen
Posts: 1742
Joined: March 15th, 2016, 6:41 pm
Location: re^jθ

Re: apgsearch v5.0

Post by Hdjensofjfnen » January 1st, 2020, 2:31 pm

gama wrote:
January 1st, 2020, 11:23 am

Code: Select all

Instruction set SSE3 detected
Pretty sure the error is in this line, as usually it gives you an alphanumeric string for the instruction set, like "l_B2yesvAaXBWz", not "SSE3".

Code: Select all

x = 5, y = 9, rule = B3-jqr/S01c2-in3
3bo$4bo$o2bo$2o2$2o$o2bo$4bo$3bo!

Code: Select all

x = 7, y = 5, rule = B3/S2-i3-y4i
4b3o$6bo$o3b3o$2o$bo!

gama
Posts: 17
Joined: October 24th, 2018, 7:14 am
Location: In the universe

Re: apgsearch v5.0

Post by gama » January 2nd, 2020, 12:45 am

Hdjensofjfnen wrote:
January 1st, 2020, 2:31 pm
gama wrote:
January 1st, 2020, 11:23 am

Code: Select all

Instruction set SSE3 detected
Pretty sure the error is in this line, as usually it gives you an alphanumeric string for the instruction set, like "l_B2yesvAaXBWz", not "SSE3".
Huh... so how do I fix this error?
EDIT: I checked on CPU-Z and my computer supports SSE2, x86_64 and also SSE3. This is weird...
Last edited by gama on January 2nd, 2020, 7:39 am, edited 1 time in total.

Code: Select all

x = 25, y = 26, rule = Crawlers
21.A2$19.A2.A.A$22.ABA$17.A2.A$20.ABA$15.A2.A$18.ABA$13.A2.A$16.ABA$
11.A2.A$14.ABA$9.A2.A$12.ABA$7.A2.A$10.ABA$5.A2.A$8.ABA$3.A2.A$6.ABA$
.A2.A$4.ABA$2.A$2.ABA2$2BA!

User avatar
dvgrn
Moderator
Posts: 10565
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: apgsearch v5.0

Post by dvgrn » January 2nd, 2020, 7:14 am

gama wrote:
January 2nd, 2020, 12:45 am
Huh... so how do I fix this error?
EDIT: I checked on CPU-Z and my computer supports SSE2 and also SSE3. This is weird...
It sounds fairly normal, if your computer is old enough. SSE3 was introduced way back in 2004 as a superset of SSE3. Are you trying to run apgsearch on a 32-bit computer? If so, see the documentation:
Note: apgluxe can only run on x86-64 machines. If you have an ancient computer, then the recommended alternative is the Python script.

gama
Posts: 17
Joined: October 24th, 2018, 7:14 am
Location: In the universe

Re: apgsearch v5.0

Post by gama » January 2nd, 2020, 9:42 am

gama wrote: I checked on CPU-Z and my computer supports SSE2, x86_64 and also SSE3. This is weird and still remains unfixed...

Code: Select all

x = 25, y = 26, rule = Crawlers
21.A2$19.A2.A.A$22.ABA$17.A2.A$20.ABA$15.A2.A$18.ABA$13.A2.A$16.ABA$
11.A2.A$14.ABA$9.A2.A$12.ABA$7.A2.A$10.ABA$5.A2.A$8.ABA$3.A2.A$6.ABA$
.A2.A$4.ABA$2.A$2.ABA2$2BA!

User avatar
muzik
Posts: 5578
Joined: January 28th, 2016, 2:47 pm
Location: Scotland

Re: apgsearch v5.0

Post by muzik » January 4th, 2020, 12:43 am

It's been a while, when will the last two hex symmetries be supported?

(There's also no demonstrations of hex symmetry on the wiki yet...)

User avatar
muzik
Posts: 5578
Joined: January 28th, 2016, 2:47 pm
Location: Scotland

Re: apgsearch v5.0

Post by muzik » January 15th, 2020, 10:11 pm

I recently submitted a patch (if that isn't enough to get Apple Bottom squealing in sheer excitement then I'm not sure what is) that introduces the ability to search gutter-symmetric soups on apgsearch. Currently implemented are D2_+1 gutters, with a skew of 0, 1 or 2. I don't know how to do diagonal gutters or skewgutters.

Currently there aren't any checks as to whether a searched rule actually preserves the selected gutter symmetry, but I'm not sure if that's a particularly major issue at all anyway since inflated soups aren't restricted to margolus-type rules.

But that said, have fun with these three new symmetries (and here's to finding a natural centinal or something)!



EDIT: HANG ON A SECOND. IT SEEMS TO BE GENERATING THESE SOUPS AS C1 INSTEAD. I don't know what's causing this but the submitted hauls will probably need to be backed out pronto. I swear I had all this completely sorted in the testing phase...

Here's an extract from the old working hashsoup:

Code: Select all

if (symmetry == "D2_xo") { return shift_bitworld(vbw, 0, -15); }

        if (symmetry == "D2_+1") { bw += shift_bitworld(vbw, 0, -15); return bw; }
        if (symmetry == "D2_+2") { bw += shift_bitworld(vbw, 0, -16); return bw; }
			// To do: only allow rules without B2(c,i),4(c,i),6(c,i) for the following as these as these transitions break gutter symmetry
		if (symmetry == "D2_+1_gO1s0_MBbeta1") { bw += shift_bitworld(vbw, 0, -17); return bw; }
			// To do: figure out the exact transitions that break the following skewgutters
		if (symmetry == "D2_+1_gO1s1_MBbeta1") { bw += shift_bitworld(vbw, 1, -17); return bw; }
		if (symmetry == "D2_+1_gO1s2_MBbeta1") { bw += shift_bitworld(vbw, 2, -17); return bw; }

        bitworld hbw;
and from the working samples.py (does the order have something to do with it?)

Code: Select all

if rulestring[-1].lower() == 'h':
        validsyms += ["C2_4", "C2_1", "C3_1", "C6", "D2_x", "D2_xo", "D4_x4", "D4_x1", "D6_1", "D6_1o", "D12"]
    else:
        validsyms += ["C2_4", "C2_2", "C2_1", "C4_4", "C4_1",
                        "D2_+2", "D2_+1", "D2_+1_gO1s0_MBbeta1", "D2_+1_gO1s1_MBbeta1", "D2_+1_gO1s2_MBbeta1", "D2_x", "D4_+4", "D4_+2", "D4_+1", "D4_x4", "D4_x1", "D8_4", "D8_1"]

    if (re.match('^b0?1?2?3?4?5?6?7?8?s0?1?2?3?4?5?6?7?8?$', rulestring)):

GUYTU6J
Posts: 2200
Joined: August 5th, 2016, 10:27 am
Location: 拆哪!I repeat, CHINA! (a.k.a. 种花家)
Contact:

Re: apgsearch v5.0

Post by GUYTU6J » February 1st, 2020, 12:33 am

I took the 5g creator here, modified const int gc to 4, compiled with g++ -o 4g 4g.cpp -O3 -Ofast -flto -march=native giving a 4g.exe. Then I ran the command in cygwin terminal:

Code: Select all

./recompile.sh --rule b3s23 --symmetry 4Glider_stdin
4g.exe | apgluxe.exe -k <KEY> -n 300000
The search was done but it gave a log text. How do I upload the results to the Catagolue b3s23/4Glider_stdin census?

wildmyron
Posts: 1542
Joined: August 9th, 2013, 12:45 am
Location: Western Australia

Re: apgsearch v5.0

Post by wildmyron » February 1st, 2020, 1:41 am

GUYTU6J wrote:
February 1st, 2020, 12:33 am
I took the 5g creator here, modified const int gc to 4, compiled with g++ -o 4g 4g.cpp -O3 -Ofast -flto -march=native giving a 4g.exe. Then I ran the command in cygwin terminal:

Code: Select all

./recompile.sh --rule b3s23 --symmetry 4Glider_stdin
4g.exe | apgluxe.exe -k <KEY> -n 300000
The search was done but it gave a log text. How do I upload the results to the Catagolue b3s23/4Glider_stdin census?
When running with stdin symmetries apgluxe uses testing mode by default. Add "-t 0" to the command to disable testing and upload to Catagolue (and "-L 0" to disable logging)
The 5S project (Smallest Spaceships Supporting Specific Speeds) is now maintained by AforAmpere. The latest collection is hosted on GitHub and contains well over 1,000,000 spaceships.

Semi-active here - recovering from a severe case of LWTDS.

User avatar
rowett
Moderator
Posts: 3765
Joined: January 31st, 2013, 2:34 am
Location: UK
Contact:

Re: apgsearch v5.0

Post by rowett » February 28th, 2020, 10:15 am

A note on performance on Windows with Zen2.

Platform:
  • Windows 10 Home 64bit
  • CPU: AMD Ryzen 9 3950X 16-core/32-thread
  • RAM: 32Gb DDR4 3600MHz
I tried three different builds:
  1. the precompiled windows binary apgluxe-windows-x86_64.exe
  2. building on Cygwin64
  3. building on WLS (Windows Subsystem for Linux)
First I tested a single thread on each different build:
soups-env.png
soups-env.png (30.48 KiB) Viewed 15459 times
The native build is clearly not optimized for the Zen 2 architecture. The WLS build is 8% faster than the Cygwin64 build.

Using the WLS build I then ran 32 searches from 1 thread to 32 threads using:

Code: Select all

% ./apgluxe -n 10000000 -p <threads>
soups-total.png
soups-total.png (43.97 KiB) Viewed 15459 times
With all 32 threads I got a peak of 167,757 soups per second.

The CPU average frequency reduces based on heat and load as the number of threads increases:
soups-cpu.png
soups-cpu.png (41.57 KiB) Viewed 15459 times
soups-temp.png
soups-temp.png (36.89 KiB) Viewed 15459 times
Here are the results as text:

Code: Select all

Threads	Soups/S	MHz	Temp
1	8398	4430	65
2	16655	4404	69
3	24617	4377	68
4	32031	4327	69
5	39717	4315	71
6	47287	4298	71
7	55003	4277	71
8	62856	4252	71
9	69992	4221	72
10	77511	4193	72
11	84787	4191	72
12	91666	4155	73
13	98664	4141	73
14	105582	4137	74
15	110734	4121	74
16	115376	4104	76
17	118479	4097	77
18	121695	4071	78
19	125090	4060	78
20	128226	4060	78
21	131114	4060	78
22	135744	4060	78
23	138243	4051	78
24	139950	4044	77
25	144310	4025	76
26	148018	4018	76
27	151203	4002	75
28	154480	3991	75
29	157678	3980	74
30	160955	3967	73
31	161449	3955	73
32	167757	3947	72
Running 4 processes with 8 threads each, and 2 processes with 16 threads each, made very little difference and was only 2% (4x8) and 1% (2x16) faster.

User avatar
gameoflifemaniac
Posts: 1242
Joined: January 22nd, 2017, 11:17 am
Location: There too

Re: apgsearch v5.0

Post by gameoflifemaniac » March 17th, 2020, 5:18 pm

I finally wanted to download apgsearch, but my computer says it may contain a virus. Can I just ignore that?
I was so socially awkward in the past and it will haunt me for the rest of my life.

Code: Select all

b4o25bo$o29bo$b3o3b3o2bob2o2bob2o2bo3bobo$4bobo3bob2o2bob2o2bobo3bobo$
4bobo3bobo5bo5bo3bobo$o3bobo3bobo5bo6b4o$b3o3b3o2bo5bo9bobo$24b4o!

User avatar
dvgrn
Moderator
Posts: 10565
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: apgsearch v5.0

Post by dvgrn » March 17th, 2020, 5:54 pm

gameoflifemaniac wrote:
March 17th, 2020, 5:18 pm
I finally wanted to download apgsearch, but my computer says it may contain a virus. Can I just ignore that?
Which file is your computer reporting as a problem, and where and how did you download it exactly?

If you're downloading the precompiled 5.09 executable from Catagolue's apgsearch page, then yes, most virus checkers these days complain bitterly about "not commonly downloaded" executables coming from the big scary Intertubes. They may even decide to quarantine the executable even after you say to keep it, in some cases, and you'll get a virus-checker popup message and the .exe file will just suddenly disappear off your hard drive.

When you download source code and compile it yourself, the resulting executable is not usually seen as scary unless it actually looks like a known virus.

User avatar
gameoflifemaniac
Posts: 1242
Joined: January 22nd, 2017, 11:17 am
Location: There too

Re: apgsearch v5.0

Post by gameoflifemaniac » March 17th, 2020, 6:13 pm

dvgrn wrote:
March 17th, 2020, 5:54 pm
gameoflifemaniac wrote:
March 17th, 2020, 5:18 pm
I finally wanted to download apgsearch, but my computer says it may contain a virus. Can I just ignore that?
Which file is your computer reporting as a problem, and where and how did you download it exactly?

If you're downloading the precompiled 5.09 executable from Catagolue's apgsearch page, then yes, most virus checkers these days complain bitterly about "not commonly downloaded" executables coming from the big scary Intertubes. They may even decide to quarantine the executable even after you say to keep it, in some cases, and you'll get a virus-checker popup message and the .exe file will just suddenly disappear off your hard drive.

When you download source code and compile it yourself, the resulting executable is not usually seen as scary unless it actually looks like a known virus.
Yes, I downloaded the pre-compiled version and luckily it didn't quarantine the executable.
I was so socially awkward in the past and it will haunt me for the rest of my life.

Code: Select all

b4o25bo$o29bo$b3o3b3o2bob2o2bob2o2bo3bobo$4bobo3bob2o2bob2o2bobo3bobo$
4bobo3bobo5bo5bo3bobo$o3bobo3bobo5bo6b4o$b3o3b3o2bo5bo9bobo$24b4o!

Post Reply