PrimeGrid
Please visit donation page to help the project cover running costs for this month

Toggle Menu

Join PrimeGrid

Returning Participants

Community

Leader Boards

Results

Other

drummers-lowrise

Advanced search

Message boards : Number crunching : Multithreading?

Author Message
Peter Hucker of the Scottish Boinc Team
Send message
Joined: 23 Sep 06
Posts: 144
ID: 3541
Credit: 68,069,390
RAC: 14,398
321 LLR Silver: Earned 100,000 credits (117,162)Cullen LLR Silver: Earned 100,000 credits (199,800)ESP LLR Silver: Earned 100,000 credits (100,301)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (100,048)PPS LLR Bronze: Earned 10,000 credits (31,226)PSP LLR Silver: Earned 100,000 credits (296,161)SoB LLR Amethyst: Earned 1,000,000 credits (1,468,570)SR5 LLR Bronze: Earned 10,000 credits (44,401)SGS LLR Silver: Earned 100,000 credits (107,261)TRP LLR Bronze: Earned 10,000 credits (62,426)Woodall LLR Silver: Earned 100,000 credits (240,222)321 Sieve (suspended) Bronze: Earned 10,000 credits (30,838)PPS Sieve Turquoise: Earned 5,000,000 credits (5,004,308)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Bronze: Earned 10,000 credits (16,965)TRP Sieve (suspended) Bronze: Earned 10,000 credits (14,492)AP 26/27 Jade: Earned 10,000,000 credits (12,275,843)GFN Sapphire: Earned 20,000,000 credits (45,717,763)WW Ruby: Earned 2,000,000 credits (2,200,000)
Message 155401 - Posted: 13 May 2022 | 2:45:01 UTC

So some apps say recommended and some say not recommended. Why are you issuing tasks in multithread when it's not recommended? I can't control this in the preferences per app. I could play around with every single app in app config, but surely you guys know that app A works best on 1 thread, app B best on 4 threads, app C best on all threads. So issue them with that number instead of letting us run our computers inefficiently.

Werinbert
Send message
Joined: 9 Jun 13
Posts: 162
ID: 233452
Credit: 361,070,764
RAC: 296,245
321 LLR Amethyst: Earned 1,000,000 credits (1,005,580)Cullen LLR Amethyst: Earned 1,000,000 credits (1,017,956)ESP LLR Amethyst: Earned 1,000,000 credits (1,010,693)Generalized Cullen/Woodall LLR Amethyst: Earned 1,000,000 credits (1,016,282)PPS LLR Jade: Earned 10,000,000 credits (10,000,222)PSP LLR Amethyst: Earned 1,000,000 credits (1,018,516)SoB LLR Gold: Earned 500,000 credits (969,575)SR5 LLR Amethyst: Earned 1,000,000 credits (1,000,484)SGS LLR Amethyst: Earned 1,000,000 credits (1,263,384)TRP LLR Amethyst: Earned 1,000,000 credits (1,004,720)Woodall LLR Gold: Earned 500,000 credits (642,276)321 Sieve (suspended) Jade: Earned 10,000,000 credits (10,014,922)Generalized Cullen/Woodall Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,101,470)PPS Sieve Double Bronze: Earned 100,000,000 credits (100,344,557)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Gold: Earned 500,000 credits (715,073)TRP Sieve (suspended) Gold: Earned 500,000 credits (630,833)AP 26/27 Emerald: Earned 50,000,000 credits (79,364,090)GFN Sapphire: Earned 20,000,000 credits (38,830,513)WW Double Bronze: Earned 100,000,000 credits (100,076,000)PSA Jade: Earned 10,000,000 credits (10,047,861)
Message 155403 - Posted: 13 May 2022 | 3:08:40 UTC - in response to Message 155401.

It is merely a recommendation not the final word. Everyone's computer is different. Everyone's crunching needs is different. One size does not fit all.


...I would have a hard time running an 8 thread task on my 4 thread computer...
____________
Werinbert is not prime... or PRPnet keeps telling me so.
Badge score: 12x3 + 1x4 + 2x6 + 2x7 + 1x8 + 1x9 + 1x10 = 93

Profile mikey
Avatar
Send message
Joined: 17 Mar 09
Posts: 1369
ID: 37043
Credit: 589,999,677
RAC: 1,887,900
Discovered 1 mega prime321 LLR Ruby: Earned 2,000,000 credits (2,066,244)Cullen LLR Ruby: Earned 2,000,000 credits (2,074,615)ESP LLR Ruby: Earned 2,000,000 credits (2,243,517)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,142,353)PPS LLR Turquoise: Earned 5,000,000 credits (7,293,199)PSP LLR Ruby: Earned 2,000,000 credits (2,232,103)SoB LLR Ruby: Earned 2,000,000 credits (2,934,612)SR5 LLR Ruby: Earned 2,000,000 credits (2,071,823)SGS LLR Turquoise: Earned 5,000,000 credits (6,780,752)TRP LLR Ruby: Earned 2,000,000 credits (2,654,806)Woodall LLR Ruby: Earned 2,000,000 credits (2,014,811)321 Sieve (suspended) Sapphire: Earned 20,000,000 credits (23,770,672)Cullen/Woodall Sieve (suspended) Gold: Earned 500,000 credits (944,431)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,813,253)PPS Sieve Double Silver: Earned 200,000,000 credits (379,305,001)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,446,797)AP 26/27 Sapphire: Earned 20,000,000 credits (33,140,471)GFN Emerald: Earned 50,000,000 credits (66,527,214)WW Turquoise: Earned 5,000,000 credits (8,092,000)PSA Sapphire: Earned 20,000,000 credits (20,457,430)
Message 155404 - Posted: 13 May 2022 | 3:31:14 UTC - in response to Message 155401.

So some apps say recommended and some say not recommended. Why are you issuing tasks in multithread when it's not recommended? I can't control this in the preferences per app. I could play around with every single app in app config, but surely you guys know that app A works best on 1 thread, app B best on 4 threads, app C best on all threads. So issue them with that number instead of letting us run our computers inefficiently.


There are like 15 different venues here at Prime Grid, put each pc in it's own venue and tweak the threads per core as needed for each one.

The problem is Intel cpu's have like 5 or different versions each with different L2 and L3 cache sizes meaning 4 treads running at the same time can be a bad thing but 3 threads running at the same time is the sweet spot for this cpu but not that one. Yes it can be a pain in the butt given all the different kind of tasks they have here and the kind of cpu's we all have.

The short answer is throw lots of cpu cores at the tasks that take a long time and throw fewer cpu cores at the ones that take less time on average. For Intel cpu's the rule of thumb is to NOT use HT as it shares the L2 and L3 cache with the main cpu core and cause slower run times. AMD cpu's are NOT the same but if you start out not using the HT cores it's not a bad thing.

Profile composite
Volunteer tester
Send message
Joined: 16 Feb 10
Posts: 1000
ID: 55391
Credit: 876,960,392
RAC: 336,036
Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2022 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (6,055,323)Cullen LLR Gold: Earned 500,000 credits (776,297)ESP LLR Ruby: Earned 2,000,000 credits (3,433,680)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,443,837)PPS LLR Sapphire: Earned 20,000,000 credits (33,578,023)PSP LLR Turquoise: Earned 5,000,000 credits (6,587,988)SoB LLR Sapphire: Earned 20,000,000 credits (44,987,733)SR5 LLR Turquoise: Earned 5,000,000 credits (6,205,694)SGS LLR Ruby: Earned 2,000,000 credits (3,627,819)TRP LLR Turquoise: Earned 5,000,000 credits (7,078,152)Woodall LLR Amethyst: Earned 1,000,000 credits (1,693,614)321 Sieve (suspended) Emerald: Earned 50,000,000 credits (50,256,050)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,571,178)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,009,610)PPS Sieve Double Silver: Earned 200,000,000 credits (455,672,175)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,165,888)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,071,454)AP 26/27 Turquoise: Earned 5,000,000 credits (6,798,063)GFN Emerald: Earned 50,000,000 credits (57,113,430)WW Ruby: Earned 2,000,000 credits (2,072,000)PSA Double Bronze: Earned 100,000,000 credits (102,762,384)
Message 155405 - Posted: 13 May 2022 | 5:11:24 UTC - in response to Message 155404.

For Intel cpu's the rule of thumb is to NOT use HT as it shares the L2 and L3 cache with the main cpu core and cause slower run times.

mikey meant to say that the pair of HT threads in each core share the core's L1 and L2 caches.

Profile DaveProject donor
Avatar
Send message
Joined: 13 Feb 12
Posts: 3042
ID: 130544
Credit: 2,058,765,226
RAC: 913,076
Found 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Jade: Earned 10,000,000 credits (10,033,365)Cullen LLR Jade: Earned 10,000,000 credits (10,122,385)ESP LLR Jade: Earned 10,000,000 credits (10,063,828)Generalized Cullen/Woodall LLR Jade: Earned 10,000,000 credits (10,095,903)PPS LLR Jade: Earned 10,000,000 credits (10,129,983)PSP LLR Jade: Earned 10,000,000 credits (10,052,227)SoB LLR Jade: Earned 10,000,000 credits (10,284,287)SR5 LLR Jade: Earned 10,000,000 credits (10,804,319)SGS LLR Jade: Earned 10,000,000 credits (10,021,480)TRP LLR Jade: Earned 10,000,000 credits (10,047,870)Woodall LLR Jade: Earned 10,000,000 credits (10,166,586)321 Sieve (suspended) Jade: Earned 10,000,000 credits (10,003,334)Cullen/Woodall Sieve (suspended) Silver: Earned 100,000 credits (268,250)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,000,502)PPS Sieve Double Silver: Earned 200,000,000 credits (310,223,017)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,000,133)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,000,970)AP 26/27 Double Silver: Earned 200,000,000 credits (206,395,150)GFN Double Silver: Earned 200,000,000 credits (204,159,637)WW Double Gold: Earned 500,000,000 credits (985,892,000)PSA Double Silver: Earned 200,000,000 credits (200,000,001)
Message 155407 - Posted: 13 May 2022 | 7:55:17 UTC

Concentrate on one project at once. Then when you change project change the MT threads as well.

Profile mikey
Avatar
Send message
Joined: 17 Mar 09
Posts: 1369
ID: 37043
Credit: 589,999,677
RAC: 1,887,900
Discovered 1 mega prime321 LLR Ruby: Earned 2,000,000 credits (2,066,244)Cullen LLR Ruby: Earned 2,000,000 credits (2,074,615)ESP LLR Ruby: Earned 2,000,000 credits (2,243,517)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,142,353)PPS LLR Turquoise: Earned 5,000,000 credits (7,293,199)PSP LLR Ruby: Earned 2,000,000 credits (2,232,103)SoB LLR Ruby: Earned 2,000,000 credits (2,934,612)SR5 LLR Ruby: Earned 2,000,000 credits (2,071,823)SGS LLR Turquoise: Earned 5,000,000 credits (6,780,752)TRP LLR Ruby: Earned 2,000,000 credits (2,654,806)Woodall LLR Ruby: Earned 2,000,000 credits (2,014,811)321 Sieve (suspended) Sapphire: Earned 20,000,000 credits (23,770,672)Cullen/Woodall Sieve (suspended) Gold: Earned 500,000 credits (944,431)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,813,253)PPS Sieve Double Silver: Earned 200,000,000 credits (379,305,001)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,446,797)AP 26/27 Sapphire: Earned 20,000,000 credits (33,140,471)GFN Emerald: Earned 50,000,000 credits (66,527,214)WW Turquoise: Earned 5,000,000 credits (8,092,000)PSA Sapphire: Earned 20,000,000 credits (20,457,430)
Message 155410 - Posted: 13 May 2022 | 12:22:10 UTC - in response to Message 155405.

For Intel cpu's the rule of thumb is to NOT use HT as it shares the L2 and L3 cache with the main cpu core and cause slower run times.


mikey meant to say that the pair of HT threads in each core share the core's L1 and L2 caches.


Yes I did, thanks for catching that.

mackerelProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Oct 08
Posts: 2583
ID: 29980
Credit: 550,189,840
RAC: 37,015
Discovered 6 mega primesEliminated 1 conjecture "k"Found 3 primes in the 2018 Tour de PrimesFound 1 mega prime in the 2018 Tour de PrimesFound 5 primes in the 2019 Tour de PrimesFound 6 primes in the 2020 Tour de PrimesFound 5 primes in the 2021 Tour de PrimesFound 1 prime in the 2022 Tour de Primes321 LLR Jade: Earned 10,000,000 credits (10,736,922)Cullen LLR Turquoise: Earned 5,000,000 credits (5,903,451)ESP LLR Turquoise: Earned 5,000,000 credits (7,207,880)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (6,714,227)PPS LLR Double Bronze: Earned 100,000,000 credits (119,496,561)PSP LLR Jade: Earned 10,000,000 credits (16,676,406)SoB LLR Sapphire: Earned 20,000,000 credits (20,019,367)SR5 LLR Sapphire: Earned 20,000,000 credits (26,030,253)SGS LLR Turquoise: Earned 5,000,000 credits (7,361,220)TPS LLR (retired) Bronze: Earned 10,000 credits (34,130)TRP LLR Sapphire: Earned 20,000,000 credits (35,623,812)Woodall LLR Turquoise: Earned 5,000,000 credits (8,171,820)321 Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,236,219)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,383,853)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,626,419)PPS Sieve Emerald: Earned 50,000,000 credits (76,969,144)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,293,882)TRP Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,012,757)AP 26/27 Sapphire: Earned 20,000,000 credits (27,813,588)GFN Emerald: Earned 50,000,000 credits (90,898,231)WW Sapphire: Earned 20,000,000 credits (34,040,000)PSA Ruby: Earned 2,000,000 credits (2,939,755)
Message 155411 - Posted: 13 May 2022 | 15:42:34 UTC

Note my following answer applies to work done by LLR.

You have to define what is optimal. For most people most of the time outside of some events, this is doing the most work in a given time. That is what I will focus on. There may be edge cases where you desire to do a single task faster but less throughput efficient.

The best throughput will mostly depend on the FFT size of the task being worked on, and the characteristics of the CPU used, specifically core count and effective cache sizes. The optimum number of tasks run simultaneously may vary due to this.

The work done uses a lot of data bandwidth. In short, if that data can be served by memory, or failing that, the CPU cache, you get the most work done.
Small work fitting within the cache per core will be best one one task per core regardless of the CPU.
To reduce the demand you can run multiple cores per task since they share data. The tradeoff here is that scaling is not perfect, and efficiency drops with more threads per task.
A rule of thumb I came up with in the past is you want to fit the most tasks as you can into the CPU cache, but not exceed it. Already it can be seen this will vary with the tasks and CPU used, so there is no fixed solution. You may choose for whatever reason not to use 100% of CPU cores, and in that case the optimisation will vary.

In case it is wondered why I refer to it as effective cache, I don't claim to fully understand it. For most CPUs it is sufficient to consider only the L3 cache size. Intel's Skylake-X and descendants seems to work better if you consider its combined L2+L3 cache. While it uses non-inclusive cache, trying to apply the same to other non-inclusive cache designs or exclusive cache designs doesn't seem to work the same.

Ryzen CPUs offer another factor to consider: CCX. What I wrote above only works for a unified cache. But Ryzen's scaling is in part achieved by breaking the CPU down into smaller logical sections. In that case, you have to consider it per CCX. Got a Ryzen 3600? It is not a 6 core CPU, it is two 3 core CPUs for the purposes of optimisation. That can lead to complications depending on how the OS scheduler works. I had a performance increase when it was recommended I run Process Lasso suck that tasks remain on a given CCX, as them migrating between them will reduce performance. On that specific CPU, running three tasks of 2 cores each is a bad idea.

In my past testing of HT, I never proved a significant gain if you compare the ideal scenario without HT and the ideal scenario with HT. There were the slightest of hints there might be 1-2% increase in throughput but this was well within the margin of error. Also, using HT does significantly increase power usage. I didn't investigate further due to that. Likewise I didn't see any different with SMT on Ryzen, although I didn't test it in as much depth.

It should be cautioned I did see gains from HT if you have a non-ideal scenario. Basically running more threads than cores can help mask losses elsewhere. It feels like Windows scheduler does better now than it used to, so performance seem closer to ideal without having to use tricks like affinity setting.

My observations were done up to Zen 2 on AMD side, and Comet Lake on Intel side. I do have Rocket Lake also, but it seems close enough I never checked it in depth. Alder Lake would probably mess things up with the P+E core mix.

Peter Hucker of the Scottish Boinc Team
Send message
Joined: 23 Sep 06
Posts: 144
ID: 3541
Credit: 68,069,390
RAC: 14,398
321 LLR Silver: Earned 100,000 credits (117,162)Cullen LLR Silver: Earned 100,000 credits (199,800)ESP LLR Silver: Earned 100,000 credits (100,301)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (100,048)PPS LLR Bronze: Earned 10,000 credits (31,226)PSP LLR Silver: Earned 100,000 credits (296,161)SoB LLR Amethyst: Earned 1,000,000 credits (1,468,570)SR5 LLR Bronze: Earned 10,000 credits (44,401)SGS LLR Silver: Earned 100,000 credits (107,261)TRP LLR Bronze: Earned 10,000 credits (62,426)Woodall LLR Silver: Earned 100,000 credits (240,222)321 Sieve (suspended) Bronze: Earned 10,000 credits (30,838)PPS Sieve Turquoise: Earned 5,000,000 credits (5,004,308)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Bronze: Earned 10,000 credits (16,965)TRP Sieve (suspended) Bronze: Earned 10,000 credits (14,492)AP 26/27 Jade: Earned 10,000,000 credits (12,275,843)GFN Sapphire: Earned 20,000,000 credits (45,717,763)WW Ruby: Earned 2,000,000 credits (2,200,000)
Message 155412 - Posted: 13 May 2022 | 16:25:40 UTC

Ok, I can see that every CPU is different. But why does it say on the preferences for some apps "multithreading is not recommended"? If that is correct for all CPUs, then the server should never give out multithreaded versions. If it is not correct for all CPUs, the notice shouldn't be there.

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 441
ID: 49572
Credit: 2,359,721,112
RAC: 1,780,073
Discovered 3 mega primesFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de PrimesFound 4 primes in the 2021 Tour de PrimesFound 7 primes in the 2022 Tour de PrimesFound 1 mega prime in the 2022 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (27,589,187)Cullen LLR Sapphire: Earned 20,000,000 credits (20,754,874)ESP LLR Sapphire: Earned 20,000,000 credits (20,453,074)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (46,657,314)PPS LLR Emerald: Earned 50,000,000 credits (61,756,838)PSP LLR Sapphire: Earned 20,000,000 credits (20,571,669)SoB LLR Emerald: Earned 50,000,000 credits (50,873,282)SR5 LLR Sapphire: Earned 20,000,000 credits (23,694,369)SGS LLR Sapphire: Earned 20,000,000 credits (21,819,843)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Sapphire: Earned 20,000,000 credits (20,376,385)321 Sieve (suspended) Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (502,006,216)GFN Double Silver: Earned 200,000,000 credits (278,213,049)WW Double Silver: Earned 200,000,000 credits (462,228,000)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 155415 - Posted: 13 May 2022 | 22:36:47 UTC - in response to Message 155412.

The easy answer is user freedom. Let the individual decide what they want to do with their rigs.

The longer answer is that on those projects, the tasks are very small (will almost always fit into the CPU cache with 1 thread/task regardless of core count, though MEGA is more iffy) and for the great majority of PG-connected computers, multithreading is actually less efficient.

Some people still run ancient low-end hardware and multithreading may be the only way they can get tasks done in a certain amount of time, and when being first mattered, it was a chance for them to increase their odds.

Also, it's all the exact same executable for all LLR2 subs. Why spend time on maintaining two very slightly different versions?
____________
Eating more cheese on Thursdays.

Peter Hucker of the Scottish Boinc Team
Send message
Joined: 23 Sep 06
Posts: 144
ID: 3541
Credit: 68,069,390
RAC: 14,398
321 LLR Silver: Earned 100,000 credits (117,162)Cullen LLR Silver: Earned 100,000 credits (199,800)ESP LLR Silver: Earned 100,000 credits (100,301)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (100,048)PPS LLR Bronze: Earned 10,000 credits (31,226)PSP LLR Silver: Earned 100,000 credits (296,161)SoB LLR Amethyst: Earned 1,000,000 credits (1,468,570)SR5 LLR Bronze: Earned 10,000 credits (44,401)SGS LLR Silver: Earned 100,000 credits (107,261)TRP LLR Bronze: Earned 10,000 credits (62,426)Woodall LLR Silver: Earned 100,000 credits (240,222)321 Sieve (suspended) Bronze: Earned 10,000 credits (30,838)PPS Sieve Turquoise: Earned 5,000,000 credits (5,004,308)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Bronze: Earned 10,000 credits (16,965)TRP Sieve (suspended) Bronze: Earned 10,000 credits (14,492)AP 26/27 Jade: Earned 10,000,000 credits (12,275,843)GFN Sapphire: Earned 20,000,000 credits (45,717,763)WW Ruby: Earned 2,000,000 credits (2,200,000)
Message 155416 - Posted: 13 May 2022 | 22:55:21 UTC - in response to Message 155415.

In that case I suggest some kind of FAQ for which processors do best on which tasks at which number of threads. I would think most people don't think about it or have the time to test everything.

And why only Thursdays?

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 441
ID: 49572
Credit: 2,359,721,112
RAC: 1,780,073
Discovered 3 mega primesFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de PrimesFound 4 primes in the 2021 Tour de PrimesFound 7 primes in the 2022 Tour de PrimesFound 1 mega prime in the 2022 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (27,589,187)Cullen LLR Sapphire: Earned 20,000,000 credits (20,754,874)ESP LLR Sapphire: Earned 20,000,000 credits (20,453,074)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (46,657,314)PPS LLR Emerald: Earned 50,000,000 credits (61,756,838)PSP LLR Sapphire: Earned 20,000,000 credits (20,571,669)SoB LLR Emerald: Earned 50,000,000 credits (50,873,282)SR5 LLR Sapphire: Earned 20,000,000 credits (23,694,369)SGS LLR Sapphire: Earned 20,000,000 credits (21,819,843)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Sapphire: Earned 20,000,000 credits (20,376,385)321 Sieve (suspended) Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (502,006,216)GFN Double Silver: Earned 200,000,000 credits (278,213,049)WW Double Silver: Earned 200,000,000 credits (462,228,000)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 155417 - Posted: 14 May 2022 | 4:51:33 UTC - in response to Message 155416.
Last modified: 14 May 2022 | 4:52:52 UTC

In that case I suggest some kind of FAQ for which processors do best on which tasks at which number of threads. I would think most people don't think about it or have the time to test everything.



I think there are already several threads on how to determine the optimal settings for one's CPU(s). I can spend a few hours working through task size vs. cache/core just on a single CPU of mine, then come back the next day after running work to check my results and the resulting steady state power/temp/utilization. While I'd have plenty of fun, personally, spending a couple months working up what would be an enormous FAQ, there are additional challenges in that most subproject tasks are increasing in FFT size (i.e. cache amount used), and so the recommendations will change possibly frequently for some CPUs but not others, which requires lots of tracking, and anyone using their integrated graphics for crunching would have to make different choices of threading.

Actually thinking about here before my bedtime, when all truly great ideas are had, I might undertake this as a personal side project just for the fun of it, though for several classes of CPU such work would be theoretical until someone with one could provide authentic data to help out (like Alder Lake P+E or newly released architectures that have a different cache paradigm).

And why only Thursdays?


If we did it every day, you wouldn't be eating more cheese on any day. ;-)
____________
Eating more cheese on Thursdays.

Peter Hucker of the Scottish Boinc Team
Send message
Joined: 23 Sep 06
Posts: 144
ID: 3541
Credit: 68,069,390
RAC: 14,398
321 LLR Silver: Earned 100,000 credits (117,162)Cullen LLR Silver: Earned 100,000 credits (199,800)ESP LLR Silver: Earned 100,000 credits (100,301)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (100,048)PPS LLR Bronze: Earned 10,000 credits (31,226)PSP LLR Silver: Earned 100,000 credits (296,161)SoB LLR Amethyst: Earned 1,000,000 credits (1,468,570)SR5 LLR Bronze: Earned 10,000 credits (44,401)SGS LLR Silver: Earned 100,000 credits (107,261)TRP LLR Bronze: Earned 10,000 credits (62,426)Woodall LLR Silver: Earned 100,000 credits (240,222)321 Sieve (suspended) Bronze: Earned 10,000 credits (30,838)PPS Sieve Turquoise: Earned 5,000,000 credits (5,004,308)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Bronze: Earned 10,000 credits (16,965)TRP Sieve (suspended) Bronze: Earned 10,000 credits (14,492)AP 26/27 Jade: Earned 10,000,000 credits (12,275,843)GFN Sapphire: Earned 20,000,000 credits (45,717,763)WW Ruby: Earned 2,000,000 credits (2,200,000)
Message 155418 - Posted: 14 May 2022 | 5:14:12 UTC - in response to Message 155417.

I think there are already several threads on how to determine the optimal settings for one's CPU(s). I can spend a few hours working through task size vs. cache/core just on a single CPU of mine, then come back the next day after running work to check my results and the resulting steady state power/temp/utilization. While I'd have plenty of fun, personally, spending a couple months working up what would be an enormous FAQ, there are additional challenges in that most subproject tasks are increasing in FFT size (i.e. cache amount used), and so the recommendations will change possibly frequently for some CPUs but not others, which requires lots of tracking, and anyone using their integrated graphics for crunching would have to make different choices of threading.

It must be possible to collate this data and provide a rough guide on the website for all to see. In fact.... surely given the cache size and the FFT size, it should be easy to calculate what would be best without having to actually try it?

If we did it every day, you wouldn't be eating more cheese on any day. ;-)

It would be more than other people, or more than last year. Damn it you just made me eat some.

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 441
ID: 49572
Credit: 2,359,721,112
RAC: 1,780,073
Discovered 3 mega primesFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de PrimesFound 4 primes in the 2021 Tour de PrimesFound 7 primes in the 2022 Tour de PrimesFound 1 mega prime in the 2022 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (27,589,187)Cullen LLR Sapphire: Earned 20,000,000 credits (20,754,874)ESP LLR Sapphire: Earned 20,000,000 credits (20,453,074)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (46,657,314)PPS LLR Emerald: Earned 50,000,000 credits (61,756,838)PSP LLR Sapphire: Earned 20,000,000 credits (20,571,669)SoB LLR Emerald: Earned 50,000,000 credits (50,873,282)SR5 LLR Sapphire: Earned 20,000,000 credits (23,694,369)SGS LLR Sapphire: Earned 20,000,000 credits (21,819,843)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Sapphire: Earned 20,000,000 credits (20,376,385)321 Sieve (suspended) Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (502,006,216)GFN Double Silver: Earned 200,000,000 credits (278,213,049)WW Double Silver: Earned 200,000,000 credits (462,228,000)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 155426 - Posted: 14 May 2022 | 18:55:21 UTC - in response to Message 155418.

I think there are already several threads on how to determine the optimal settings for one's CPU(s). I can spend a few hours working through task size vs. cache/core just on a single CPU of mine, then come back the next day after running work to check my results and the resulting steady state power/temp/utilization. While I'd have plenty of fun, personally, spending a couple months working up what would be an enormous FAQ, there are additional challenges in that most subproject tasks are increasing in FFT size (i.e. cache amount used), and so the recommendations will change possibly frequently for some CPUs but not others, which requires lots of tracking, and anyone using their integrated graphics for crunching would have to make different choices of threading.

It must be possible to collate this data and provide a rough guide on the website for all to see. In fact.... surely given the cache size and the FFT size, it should be easy to calculate what would be best without having to actually try it?


Yes, a rough guide is absolutely possible, but you've partially solved your own problem in the process: FFT sizes vs. cache. Wikipedia has an essentially complete specs list of all the major CPUs (ES/QS oddities notwithstanding, nor whatever allocation you get from cloud hosts). Look up your CPU's cache, the FFT size of what subproject you want to do (x8 to get MB/task) and do some simple division. Further optimization is then up to the individual's own testing.

I've started working up a guide as a spreadsheet, (goodness knows I'm not going to type it out in a post window!), which would allow for easy(er) updating as FFTs increase (and because I don't have admin privileges). There are literally thousands of different CPU models across vendors and generations that need to be collated into generic categories when possible. Understatement: This will take some time.

If we did it every day, you wouldn't be eating more cheese on any day. ;-)

It would be more than other people, or more than last year. Damn it you just made me eat some.


Excellent, my work here is done. Cheese is so tasty!
/It's just a silly quasisensical catchphrase I made up 25 years ago in high school.
____________
Eating more cheese on Thursdays.

Peter Hucker of the Scottish Boinc Team
Send message
Joined: 23 Sep 06
Posts: 144
ID: 3541
Credit: 68,069,390
RAC: 14,398
321 LLR Silver: Earned 100,000 credits (117,162)Cullen LLR Silver: Earned 100,000 credits (199,800)ESP LLR Silver: Earned 100,000 credits (100,301)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (100,048)PPS LLR Bronze: Earned 10,000 credits (31,226)PSP LLR Silver: Earned 100,000 credits (296,161)SoB LLR Amethyst: Earned 1,000,000 credits (1,468,570)SR5 LLR Bronze: Earned 10,000 credits (44,401)SGS LLR Silver: Earned 100,000 credits (107,261)TRP LLR Bronze: Earned 10,000 credits (62,426)Woodall LLR Silver: Earned 100,000 credits (240,222)321 Sieve (suspended) Bronze: Earned 10,000 credits (30,838)PPS Sieve Turquoise: Earned 5,000,000 credits (5,004,308)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Bronze: Earned 10,000 credits (16,965)TRP Sieve (suspended) Bronze: Earned 10,000 credits (14,492)AP 26/27 Jade: Earned 10,000,000 credits (12,275,843)GFN Sapphire: Earned 20,000,000 credits (45,717,763)WW Ruby: Earned 2,000,000 credits (2,200,000)
Message 155429 - Posted: 14 May 2022 | 21:08:42 UTC - in response to Message 155426.
Last modified: 14 May 2022 | 21:09:15 UTC

Yes, a rough guide is absolutely possible, but you've partially solved your own problem in the process: FFT sizes vs. cache. Wikipedia has an essentially complete specs list of all the major CPUs (ES/QS oddities notwithstanding, nor whatever allocation you get from cloud hosts). Look up your CPU's cache, the FFT size of what subproject you want to do (x8 to get MB/task) and do some simple division. Further optimization is then up to the individual's own testing.
I just think most people won't bother with all that testing or just assume the server will give them what's most sensible. I was running max cores for years until I saw someone mention in the forum that isn't efficient. I was surprised recently to change 16 cores per task to 4 cores per task and get the same throughput exactly on my CPU per task, so 4 times the work altogether.

I've started working up a guide as a spreadsheet, (goodness knows I'm not going to type it out in a post window!), which would allow for easy(er) updating as FFTs increase (and because I don't have admin privileges). There are literally thousands of different CPU models across vendors and generations that need to be collated into generic categories when possible. Understatement: This will take some time.
Good on you. I hope it's placed somewhere prominent so people have a good idea of what to run. I'd help, but I've had enough of making spreadsheets, just plastered about trying to solve a Rosetta compatibility problem.

Excellent, my work here is done. Cheese is so tasty!
/It's just a silly quasisensical catchphrase I made up 25 years ago in high school.
You're younger than me then. I've got cheddar (the double price posh stuff), stilton, wensleydale, lancashire, and greek here. The lancashire actually says on it "proper tasty lancashire" - I think you have to say that in the accent of Brian Johnson from AC DC:https://en.wikipedia.org/wiki/Brian_Johnson

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 441
ID: 49572
Credit: 2,359,721,112
RAC: 1,780,073
Discovered 3 mega primesFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de PrimesFound 4 primes in the 2021 Tour de PrimesFound 7 primes in the 2022 Tour de PrimesFound 1 mega prime in the 2022 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (27,589,187)Cullen LLR Sapphire: Earned 20,000,000 credits (20,754,874)ESP LLR Sapphire: Earned 20,000,000 credits (20,453,074)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (46,657,314)PPS LLR Emerald: Earned 50,000,000 credits (61,756,838)PSP LLR Sapphire: Earned 20,000,000 credits (20,571,669)SoB LLR Emerald: Earned 50,000,000 credits (50,873,282)SR5 LLR Sapphire: Earned 20,000,000 credits (23,694,369)SGS LLR Sapphire: Earned 20,000,000 credits (21,819,843)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Sapphire: Earned 20,000,000 credits (20,376,385)321 Sieve (suspended) Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (502,006,216)GFN Double Silver: Earned 200,000,000 credits (278,213,049)WW Double Silver: Earned 200,000,000 credits (462,228,000)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 155432 - Posted: 14 May 2022 | 23:05:02 UTC

I'm making surprisingly good progress so far, at least with Intel desktop CPUs and servers starting with Sandy Bridge. Mobile is a mess to sort out later (as is Alder Lake) and I'm purposely leaving out the low end stuff since it doesn't show up on the CPU performance chart. Peter, when I'm ready to share, I'll start a new thread that might get stickied.

Question for those in the know (programming team?): How is cache allocation/task size calculated or viewed for GFN, AP27 and WW? I didn't see anything useful in stderr's. Or is it that these apps are really suited for GPUs and as such are so massively parallel that you just want to assign all cores+logical threads, anyway?
____________
Eating more cheese on Thursdays.

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 441
ID: 49572
Credit: 2,359,721,112
RAC: 1,780,073
Discovered 3 mega primesFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de PrimesFound 4 primes in the 2021 Tour de PrimesFound 7 primes in the 2022 Tour de PrimesFound 1 mega prime in the 2022 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (27,589,187)Cullen LLR Sapphire: Earned 20,000,000 credits (20,754,874)ESP LLR Sapphire: Earned 20,000,000 credits (20,453,074)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (46,657,314)PPS LLR Emerald: Earned 50,000,000 credits (61,756,838)PSP LLR Sapphire: Earned 20,000,000 credits (20,571,669)SoB LLR Emerald: Earned 50,000,000 credits (50,873,282)SR5 LLR Sapphire: Earned 20,000,000 credits (23,694,369)SGS LLR Sapphire: Earned 20,000,000 credits (21,819,843)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Sapphire: Earned 20,000,000 credits (20,376,385)321 Sieve (suspended) Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (502,006,216)GFN Double Silver: Earned 200,000,000 credits (278,213,049)WW Double Silver: Earned 200,000,000 credits (462,228,000)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 155433 - Posted: 15 May 2022 | 1:11:07 UTC - in response to Message 155432.

I've been looking at getting now-cheap used Broadwell-EP CPUs for my X99s and really getting into the cores x frequency as a measure of general overall performance when I'm looking at different chips for roughly the same price. One of the fun things I'm learning as I make this chart is that for many-cored CPUs, the more factors that the number of cores has, the better result you're likely going to have with crunching if you want to fully utilize your CPU (and for my part, if I wasn't going to use all the cores, I'd just buy a cheaper chip with fewer cores).

For example, I'm currently working on the Icelake-SP table and a 26-core has choices of 1, 2, 13 and 26 threads, but the 24-core has 1, 2, 3, 4, 6, 8, 12, and 24 which much more nicely fits the FFT distribution. On 321, the 26-core would have to run 2x13 threads using just 18/71.5MB of cache vs. 6x4 threads using 54/66MB of cache for the 24-core. In my own experience with the variety of core-count systems I have, keeping the number of threads as low as possible is the most efficient way to go, so that 26-core would likely end up having much lower overall throughput on the non-tiny subprojects.
____________
Eating more cheese on Thursdays.

Peter Hucker of the Scottish Boinc Team
Send message
Joined: 23 Sep 06
Posts: 144
ID: 3541
Credit: 68,069,390
RAC: 14,398
321 LLR Silver: Earned 100,000 credits (117,162)Cullen LLR Silver: Earned 100,000 credits (199,800)ESP LLR Silver: Earned 100,000 credits (100,301)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (100,048)PPS LLR Bronze: Earned 10,000 credits (31,226)PSP LLR Silver: Earned 100,000 credits (296,161)SoB LLR Amethyst: Earned 1,000,000 credits (1,468,570)SR5 LLR Bronze: Earned 10,000 credits (44,401)SGS LLR Silver: Earned 100,000 credits (107,261)TRP LLR Bronze: Earned 10,000 credits (62,426)Woodall LLR Silver: Earned 100,000 credits (240,222)321 Sieve (suspended) Bronze: Earned 10,000 credits (30,838)PPS Sieve Turquoise: Earned 5,000,000 credits (5,004,308)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Bronze: Earned 10,000 credits (16,965)TRP Sieve (suspended) Bronze: Earned 10,000 credits (14,492)AP 26/27 Jade: Earned 10,000,000 credits (12,275,843)GFN Sapphire: Earned 20,000,000 credits (45,717,763)WW Ruby: Earned 2,000,000 credits (2,200,000)
Message 155434 - Posted: 15 May 2022 | 2:02:51 UTC - in response to Message 155432.
Last modified: 15 May 2022 | 2:03:01 UTC

I'm making surprisingly good progress so far, at least with Intel desktop CPUs and servers starting with Sandy Bridge. Mobile is a mess to sort out later (as is Alder Lake) and I'm purposely leaving out the low end stuff since it doesn't show up on the CPU performance chart. Peter, when I'm ready to share, I'll start a new thread that might get stickied.

That still won't reach enough people, it needs to be available when you're choosing projects.

I'm not currently in a good mood. I got a new card (different model to the others) and connected it to a machine with the other kind. Windows saw it, Boinc didn't (typical). I assumed this was because Windows hadn't put a compute enabled driver in, so I installed the latest driver for it. This corrupted the registry (what?!). So half my equipment ain't crunching for a bit.

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 441
ID: 49572
Credit: 2,359,721,112
RAC: 1,780,073
Discovered 3 mega primesFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de PrimesFound 4 primes in the 2021 Tour de PrimesFound 7 primes in the 2022 Tour de PrimesFound 1 mega prime in the 2022 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (27,589,187)Cullen LLR Sapphire: Earned 20,000,000 credits (20,754,874)ESP LLR Sapphire: Earned 20,000,000 credits (20,453,074)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (46,657,314)PPS LLR Emerald: Earned 50,000,000 credits (61,756,838)PSP LLR Sapphire: Earned 20,000,000 credits (20,571,669)SoB LLR Emerald: Earned 50,000,000 credits (50,873,282)SR5 LLR Sapphire: Earned 20,000,000 credits (23,694,369)SGS LLR Sapphire: Earned 20,000,000 credits (21,819,843)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Sapphire: Earned 20,000,000 credits (20,376,385)321 Sieve (suspended) Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (502,006,216)GFN Double Silver: Earned 200,000,000 credits (278,213,049)WW Double Silver: Earned 200,000,000 credits (462,228,000)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 155436 - Posted: 15 May 2022 | 4:12:35 UTC - in response to Message 155434.

I'm making surprisingly good progress so far, at least with Intel desktop CPUs and servers starting with Sandy Bridge. Mobile is a mess to sort out later (as is Alder Lake) and I'm purposely leaving out the low end stuff since it doesn't show up on the CPU performance chart. Peter, when I'm ready to share, I'll start a new thread that might get stickied.

That still won't reach enough people, it needs to be available when you're choosing projects.


I'm not an admin; just a nerd with a math degree, outdated programming skills, no life, and lots of free time. I'm at 400 rows of spreadsheet fun, and it still requires a user to be able to look at their CPU model and match it to a line that is generic across a specific generation like "i5 6000/7000." Things get more involved if you have a Xeon.
____________
Eating more cheese on Thursdays.

Peter Hucker of the Scottish Boinc Team
Send message
Joined: 23 Sep 06
Posts: 144
ID: 3541
Credit: 68,069,390
RAC: 14,398
321 LLR Silver: Earned 100,000 credits (117,162)Cullen LLR Silver: Earned 100,000 credits (199,800)ESP LLR Silver: Earned 100,000 credits (100,301)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (100,048)PPS LLR Bronze: Earned 10,000 credits (31,226)PSP LLR Silver: Earned 100,000 credits (296,161)SoB LLR Amethyst: Earned 1,000,000 credits (1,468,570)SR5 LLR Bronze: Earned 10,000 credits (44,401)SGS LLR Silver: Earned 100,000 credits (107,261)TRP LLR Bronze: Earned 10,000 credits (62,426)Woodall LLR Silver: Earned 100,000 credits (240,222)321 Sieve (suspended) Bronze: Earned 10,000 credits (30,838)PPS Sieve Turquoise: Earned 5,000,000 credits (5,004,308)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Bronze: Earned 10,000 credits (16,965)TRP Sieve (suspended) Bronze: Earned 10,000 credits (14,492)AP 26/27 Jade: Earned 10,000,000 credits (12,275,843)GFN Sapphire: Earned 20,000,000 credits (45,717,763)WW Ruby: Earned 2,000,000 credits (2,200,000)
Message 155437 - Posted: 15 May 2022 | 4:37:05 UTC - in response to Message 155436.

I'm making surprisingly good progress so far, at least with Intel desktop CPUs and servers starting with Sandy Bridge. Mobile is a mess to sort out later (as is Alder Lake) and I'm purposely leaving out the low end stuff since it doesn't show up on the CPU performance chart. Peter, when I'm ready to share, I'll start a new thread that might get stickied.

That still won't reach enough people, it needs to be available when you're choosing projects.


I'm not an admin; just a nerd with a math degree, outdated programming skills, no life, and lots of free time. I'm at 400 rows of spreadsheet fun, and it still requires a user to be able to look at their CPU model and match it to a line that is generic across a specific generation like "i5 6000/7000." Things get more involved if you have a Xeon.
Produce something useful and I'm sure they'll make it prominent somewhere. After all the admins do want efficiency right?

mackerelProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Oct 08
Posts: 2583
ID: 29980
Credit: 550,189,840
RAC: 37,015
Discovered 6 mega primesEliminated 1 conjecture "k"Found 3 primes in the 2018 Tour de PrimesFound 1 mega prime in the 2018 Tour de PrimesFound 5 primes in the 2019 Tour de PrimesFound 6 primes in the 2020 Tour de PrimesFound 5 primes in the 2021 Tour de PrimesFound 1 prime in the 2022 Tour de Primes321 LLR Jade: Earned 10,000,000 credits (10,736,922)Cullen LLR Turquoise: Earned 5,000,000 credits (5,903,451)ESP LLR Turquoise: Earned 5,000,000 credits (7,207,880)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (6,714,227)PPS LLR Double Bronze: Earned 100,000,000 credits (119,496,561)PSP LLR Jade: Earned 10,000,000 credits (16,676,406)SoB LLR Sapphire: Earned 20,000,000 credits (20,019,367)SR5 LLR Sapphire: Earned 20,000,000 credits (26,030,253)SGS LLR Turquoise: Earned 5,000,000 credits (7,361,220)TPS LLR (retired) Bronze: Earned 10,000 credits (34,130)TRP LLR Sapphire: Earned 20,000,000 credits (35,623,812)Woodall LLR Turquoise: Earned 5,000,000 credits (8,171,820)321 Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,236,219)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,383,853)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,626,419)PPS Sieve Emerald: Earned 50,000,000 credits (76,969,144)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,293,882)TRP Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,012,757)AP 26/27 Sapphire: Earned 20,000,000 credits (27,813,588)GFN Emerald: Earned 50,000,000 credits (90,898,231)WW Sapphire: Earned 20,000,000 credits (34,040,000)PSA Ruby: Earned 2,000,000 credits (2,939,755)
Message 155441 - Posted: 15 May 2022 | 8:31:51 UTC - in response to Message 155433.

really getting into the cores x frequency as a measure of general overall performance when I'm looking at different chips for roughly the same price.

Not forgetting architecture makes a significant contribution I hope. For example, Skylake family is about 14% ahead of Broadwell per clock. Zen 2 was about 4% up on Skylake, and Zen 3 another 10% or so on top of that. This applies to LLR like workloads when NOT limited by ram performance.

I'm currently working on the Icelake-SP table and a 26-core has choices of 1, 2, 13 and 26 threads, but the 24-core has 1, 2, 3, 4, 6, 8, 12, and 24 which much more nicely fits the FFT distribution.

At the cost of more work for you to do, there are two more scenarios you can consider:
1, not using all the cores e.g. only 24 cores out of 26. Active cores get more power budget than if all were in use so may clock slightly higher depending on CPU.
2, using more threads than cores e.g. 7 tasks of 4 threads each for 28 threads on 26 core CPU. Guarantees all cores in use.

Profile mikey
Avatar
Send message
Joined: 17 Mar 09
Posts: 1369
ID: 37043
Credit: 589,999,677
RAC: 1,887,900
Discovered 1 mega prime321 LLR Ruby: Earned 2,000,000 credits (2,066,244)Cullen LLR Ruby: Earned 2,000,000 credits (2,074,615)ESP LLR Ruby: Earned 2,000,000 credits (2,243,517)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,142,353)PPS LLR Turquoise: Earned 5,000,000 credits (7,293,199)PSP LLR Ruby: Earned 2,000,000 credits (2,232,103)SoB LLR Ruby: Earned 2,000,000 credits (2,934,612)SR5 LLR Ruby: Earned 2,000,000 credits (2,071,823)SGS LLR Turquoise: Earned 5,000,000 credits (6,780,752)TRP LLR Ruby: Earned 2,000,000 credits (2,654,806)Woodall LLR Ruby: Earned 2,000,000 credits (2,014,811)321 Sieve (suspended) Sapphire: Earned 20,000,000 credits (23,770,672)Cullen/Woodall Sieve (suspended) Gold: Earned 500,000 credits (944,431)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,813,253)PPS Sieve Double Silver: Earned 200,000,000 credits (379,305,001)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,446,797)AP 26/27 Sapphire: Earned 20,000,000 credits (33,140,471)GFN Emerald: Earned 50,000,000 credits (66,527,214)WW Turquoise: Earned 5,000,000 credits (8,092,000)PSA Sapphire: Earned 20,000,000 credits (20,457,430)
Message 155443 - Posted: 15 May 2022 | 11:43:43 UTC - in response to Message 155432.

I'm making surprisingly good progress so far, at least with Intel desktop CPUs and servers starting with Sandy Bridge. Mobile is a mess to sort out later (as is Alder Lake) and I'm purposely leaving out the low end stuff since it doesn't show up on the CPU performance chart. Peter, when I'm ready to share, I'll start a new thread that might get stickied.


I think this will help ALOT of people who are just lost when it comes to this kind of stuff, faster crunching is better crunching as long as it still validates.

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 441
ID: 49572
Credit: 2,359,721,112
RAC: 1,780,073
Discovered 3 mega primesFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de PrimesFound 4 primes in the 2021 Tour de PrimesFound 7 primes in the 2022 Tour de PrimesFound 1 mega prime in the 2022 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (27,589,187)Cullen LLR Sapphire: Earned 20,000,000 credits (20,754,874)ESP LLR Sapphire: Earned 20,000,000 credits (20,453,074)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (46,657,314)PPS LLR Emerald: Earned 50,000,000 credits (61,756,838)PSP LLR Sapphire: Earned 20,000,000 credits (20,571,669)SoB LLR Emerald: Earned 50,000,000 credits (50,873,282)SR5 LLR Sapphire: Earned 20,000,000 credits (23,694,369)SGS LLR Sapphire: Earned 20,000,000 credits (21,819,843)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Sapphire: Earned 20,000,000 credits (20,376,385)321 Sieve (suspended) Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (502,006,216)GFN Double Silver: Earned 200,000,000 credits (278,213,049)WW Double Silver: Earned 200,000,000 credits (462,228,000)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 155448 - Posted: 15 May 2022 | 17:34:10 UTC - in response to Message 155441.

really getting into the cores x frequency as a measure of general overall performance when I'm looking at different chips for roughly the same price.

Not forgetting architecture makes a significant contribution I hope. For example, Skylake family is about 14% ahead of Broadwell per clock. Zen 2 was about 4% up on Skylake, and Zen 3 another 10% or so on top of that. This applies to LLR like workloads when NOT limited by ram performance.


I'm looking to get used Xeons to upgrade the X99 boards I already have. Broadwell is the latest architecture to hit the datacenter cheap CPU sell-off, but in some cases, moving up 2 cores or adding a few hundred MHz adds $100 to the price, so going to a weird number only to use fewer cores is a waste of money, especially as once you get to about half the cores in use, the all-core lowest turbo bin is in play, so there is no frequency advantage. Cheap Skylake-SP is coming, but the disadvantages lie in needing an expensive server/workstation board and a 6000 series part is the minimum for AVX-512 to be useful.

I'm currently working on the Icelake-SP table and a 26-core has choices of 1, 2, 13 and 26 threads, but the 24-core has 1, 2, 3, 4, 6, 8, 12, and 24 which much more nicely fits the FFT distribution.

At the cost of more work for you to do, there are two more scenarios you can consider:
1, not using all the cores e.g. only 24 cores out of 26. Active cores get more power budget than if all were in use so may clock slightly higher depending on CPU.
2, using more threads than cores e.g. 7 tasks of 4 threads each for 28 threads on 26 core CPU. Guarantees all cores in use.


Unfortunately, working out different possibilities for CPUs by using fewer cores than available is well beyond the scope of the table. When it comes to the very high core count server CPUs, the table eschews model numbers and goes right to "24-core," "26-core", etc. with their common cache quantity. So if one were so inclined, while it might not be a perfect fit, could move up or down to a corresponding row to get a basic idea. It's really a starting point more than anything else, and someone who wants to tweak will work from there and explore the many other posts with real-world results.

Speaking of real-world results, I do want to thank you for all your Ryzen work. The CCX/CCD testing you've done has been a great help on the AMD side of the chart.


____________
Eating more cheese on Thursdays.

Peter Hucker of the Scottish Boinc Team
Send message
Joined: 23 Sep 06
Posts: 144
ID: 3541
Credit: 68,069,390
RAC: 14,398
321 LLR Silver: Earned 100,000 credits (117,162)Cullen LLR Silver: Earned 100,000 credits (199,800)ESP LLR Silver: Earned 100,000 credits (100,301)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (100,048)PPS LLR Bronze: Earned 10,000 credits (31,226)PSP LLR Silver: Earned 100,000 credits (296,161)SoB LLR Amethyst: Earned 1,000,000 credits (1,468,570)SR5 LLR Bronze: Earned 10,000 credits (44,401)SGS LLR Silver: Earned 100,000 credits (107,261)TRP LLR Bronze: Earned 10,000 credits (62,426)Woodall LLR Silver: Earned 100,000 credits (240,222)321 Sieve (suspended) Bronze: Earned 10,000 credits (30,838)PPS Sieve Turquoise: Earned 5,000,000 credits (5,004,308)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Bronze: Earned 10,000 credits (16,965)TRP Sieve (suspended) Bronze: Earned 10,000 credits (14,492)AP 26/27 Jade: Earned 10,000,000 credits (12,275,843)GFN Sapphire: Earned 20,000,000 credits (45,717,763)WW Ruby: Earned 2,000,000 credits (2,200,000)
Message 155453 - Posted: 15 May 2022 | 22:50:53 UTC - in response to Message 155441.
Last modified: 15 May 2022 | 22:51:10 UTC

Not forgetting architecture makes a significant contribution I hope. For example, Skylake family is about 14% ahead of Broadwell per clock. Zen 2 was about 4% up on Skylake, and Zen 3 another 10% or so on top of that. This applies to LLR like workloads when NOT limited by ram performance.

I've heard about that but never observed it, I find CPUs overclock by x amount whether you're using all the cores or not. Maybe you have to have insufficient cooling for it to slow down?

At the cost of more work for you to do, there are two more scenarios you can consider:
1, not using all the cores e.g. only 24 cores out of 26. Active cores get more power budget than if all were in use so may clock slightly higher depending on CPU.
2, using more threads than cores e.g. 7 tasks of 4 threads each for 28 threads on 26 core CPU. Guarantees all cores in use.

How on earth did you persuade Boinc to use more threads than you have? I've needed to do that on occasion and couldn't figure out a way how.

mackerelProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Oct 08
Posts: 2583
ID: 29980
Credit: 550,189,840
RAC: 37,015
Discovered 6 mega primesEliminated 1 conjecture "k"Found 3 primes in the 2018 Tour de PrimesFound 1 mega prime in the 2018 Tour de PrimesFound 5 primes in the 2019 Tour de PrimesFound 6 primes in the 2020 Tour de PrimesFound 5 primes in the 2021 Tour de PrimesFound 1 prime in the 2022 Tour de Primes321 LLR Jade: Earned 10,000,000 credits (10,736,922)Cullen LLR Turquoise: Earned 5,000,000 credits (5,903,451)ESP LLR Turquoise: Earned 5,000,000 credits (7,207,880)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (6,714,227)PPS LLR Double Bronze: Earned 100,000,000 credits (119,496,561)PSP LLR Jade: Earned 10,000,000 credits (16,676,406)SoB LLR Sapphire: Earned 20,000,000 credits (20,019,367)SR5 LLR Sapphire: Earned 20,000,000 credits (26,030,253)SGS LLR Turquoise: Earned 5,000,000 credits (7,361,220)TPS LLR (retired) Bronze: Earned 10,000 credits (34,130)TRP LLR Sapphire: Earned 20,000,000 credits (35,623,812)Woodall LLR Turquoise: Earned 5,000,000 credits (8,171,820)321 Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,236,219)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,383,853)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,626,419)PPS Sieve Emerald: Earned 50,000,000 credits (76,969,144)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,293,882)TRP Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,012,757)AP 26/27 Sapphire: Earned 20,000,000 credits (27,813,588)GFN Emerald: Earned 50,000,000 credits (90,898,231)WW Sapphire: Earned 20,000,000 credits (34,040,000)PSA Ruby: Earned 2,000,000 credits (2,939,755)
Message 155464 - Posted: 17 May 2022 | 17:11:34 UTC - in response to Message 155453.

Not forgetting architecture makes a significant contribution I hope. For example, Skylake family is about 14% ahead of Broadwell per clock. Zen 2 was about 4% up on Skylake, and Zen 3 another 10% or so on top of that. This applies to LLR like workloads when NOT limited by ram performance.

I've heard about that but never observed it, I find CPUs overclock by x amount whether you're using all the cores or not. Maybe you have to have insufficient cooling for it to slow down?

I don't understand that reply in the context quoted, since what you replied to is not affected at all by clocks.

The reply could make sense if it was referring to using fewer cores than available, but it does depend on both the CPU model and how many cores you're using. Note default turbo is not overclocking. Ryzen since Zen+ takes a more fine grain approach to turbo so you might see it more on that side.

How on earth did you persuade Boinc to use more threads than you have? I've needed to do that on occasion and couldn't figure out a way how.

I'm not? Keep in mind for most (but not all) systems you typically have 2x the threads available on CPU compared to cores through HT/SMT. There may be instances where using more threads than cores could help.

Profile composite
Volunteer tester
Send message
Joined: 16 Feb 10
Posts: 1000
ID: 55391
Credit: 876,960,392
RAC: 336,036
Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2022 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (6,055,323)Cullen LLR Gold: Earned 500,000 credits (776,297)ESP LLR Ruby: Earned 2,000,000 credits (3,433,680)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,443,837)PPS LLR Sapphire: Earned 20,000,000 credits (33,578,023)PSP LLR Turquoise: Earned 5,000,000 credits (6,587,988)SoB LLR Sapphire: Earned 20,000,000 credits (44,987,733)SR5 LLR Turquoise: Earned 5,000,000 credits (6,205,694)SGS LLR Ruby: Earned 2,000,000 credits (3,627,819)TRP LLR Turquoise: Earned 5,000,000 credits (7,078,152)Woodall LLR Amethyst: Earned 1,000,000 credits (1,693,614)321 Sieve (suspended) Emerald: Earned 50,000,000 credits (50,256,050)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,571,178)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,009,610)PPS Sieve Double Silver: Earned 200,000,000 credits (455,672,175)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,165,888)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,071,454)AP 26/27 Turquoise: Earned 5,000,000 credits (6,798,063)GFN Emerald: Earned 50,000,000 credits (57,113,430)WW Ruby: Earned 2,000,000 credits (2,072,000)PSA Double Bronze: Earned 100,000,000 credits (102,762,384)
Message 155465 - Posted: 17 May 2022 | 17:26:51 UTC

How on earth did you persuade Boinc to use more threads than you have?

You could fool BOINC into thinking that the app uses less than a full core by setting up app_config.xml
There are plenty of postings around here on how to do that.
I haven't done that lately, and I only used it to run multiple GPU tasks simultaneously.

Peter Hucker of the Scottish Boinc Team
Send message
Joined: 23 Sep 06
Posts: 144
ID: 3541
Credit: 68,069,390
RAC: 14,398
321 LLR Silver: Earned 100,000 credits (117,162)Cullen LLR Silver: Earned 100,000 credits (199,800)ESP LLR Silver: Earned 100,000 credits (100,301)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (100,048)PPS LLR Bronze: Earned 10,000 credits (31,226)PSP LLR Silver: Earned 100,000 credits (296,161)SoB LLR Amethyst: Earned 1,000,000 credits (1,468,570)SR5 LLR Bronze: Earned 10,000 credits (44,401)SGS LLR Silver: Earned 100,000 credits (107,261)TRP LLR Bronze: Earned 10,000 credits (62,426)Woodall LLR Silver: Earned 100,000 credits (240,222)321 Sieve (suspended) Bronze: Earned 10,000 credits (30,838)PPS Sieve Turquoise: Earned 5,000,000 credits (5,004,308)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Bronze: Earned 10,000 credits (16,965)TRP Sieve (suspended) Bronze: Earned 10,000 credits (14,492)AP 26/27 Jade: Earned 10,000,000 credits (12,275,843)GFN Sapphire: Earned 20,000,000 credits (45,717,763)WW Ruby: Earned 2,000,000 credits (2,200,000)
Message 155475 - Posted: 18 May 2022 | 13:09:50 UTC - in response to Message 155464.

Not forgetting architecture makes a significant contribution I hope. For example, Skylake family is about 14% ahead of Broadwell per clock. Zen 2 was about 4% up on Skylake, and Zen 3 another 10% or so on top of that. This applies to LLR like workloads when NOT limited by ram performance.

I've heard about that but never observed it, I find CPUs overclock by x amount whether you're using all the cores or not. Maybe you have to have insufficient cooling for it to slow down?

I don't understand that reply in the context quoted, since what you replied to is not affected at all by clocks.

The reply could make sense if it was referring to using fewer cores than available, but it does depend on both the CPU model and how many cores you're using. Note default turbo is not overclocking. Ryzen since Zen+ takes a more fine grain approach to turbo so you might see it more on that side.

I remember writing that in reply to something else. These project forums are running so slow just now due to the competitions, it probably messed up somehow.

How on earth did you persuade Boinc to use more threads than you have? I've needed to do that on occasion and couldn't figure out a way how.

I'm not? Keep in mind for most (but not all) systems you typically have 2x the threads available on CPU compared to cores through HT/SMT. There may be instances where using more threads than cores could help.

Sorry, I got confused with threads and cores, I just ignore that notation and consider them all "cores", since that's how they appear to the OS and any program.

Peter Hucker of the Scottish Boinc Team
Send message
Joined: 23 Sep 06
Posts: 144
ID: 3541
Credit: 68,069,390
RAC: 14,398
321 LLR Silver: Earned 100,000 credits (117,162)Cullen LLR Silver: Earned 100,000 credits (199,800)ESP LLR Silver: Earned 100,000 credits (100,301)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (100,048)PPS LLR Bronze: Earned 10,000 credits (31,226)PSP LLR Silver: Earned 100,000 credits (296,161)SoB LLR Amethyst: Earned 1,000,000 credits (1,468,570)SR5 LLR Bronze: Earned 10,000 credits (44,401)SGS LLR Silver: Earned 100,000 credits (107,261)TRP LLR Bronze: Earned 10,000 credits (62,426)Woodall LLR Silver: Earned 100,000 credits (240,222)321 Sieve (suspended) Bronze: Earned 10,000 credits (30,838)PPS Sieve Turquoise: Earned 5,000,000 credits (5,004,308)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Bronze: Earned 10,000 credits (16,965)TRP Sieve (suspended) Bronze: Earned 10,000 credits (14,492)AP 26/27 Jade: Earned 10,000,000 credits (12,275,843)GFN Sapphire: Earned 20,000,000 credits (45,717,763)WW Ruby: Earned 2,000,000 credits (2,200,000)
Message 155476 - Posted: 18 May 2022 | 13:11:53 UTC - in response to Message 155465.

How on earth did you persuade Boinc to use more threads than you have?

You could fool BOINC into thinking that the app uses less than a full core by setting up app_config.xml
There are plenty of postings around here on how to do that.
I haven't done that lately, and I only used it to run multiple GPU tasks simultaneously.

Is that by using something like I did here?

<app_version> <app_name>milkyway_nbody</app_name> <plan_class>mt</plan_class> <avg_ncpus>4</avg_ncpus> <cmdline>--nthreads 12</cmdline> </app_version>

The app can get 12 cores, but since it uses an average of 4, I tell boinc this, so it runs more of them.

Profile composite
Volunteer tester
Send message
Joined: 16 Feb 10
Posts: 1000
ID: 55391
Credit: 876,960,392
RAC: 336,036
Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2022 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (6,055,323)Cullen LLR Gold: Earned 500,000 credits (776,297)ESP LLR Ruby: Earned 2,000,000 credits (3,433,680)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,443,837)PPS LLR Sapphire: Earned 20,000,000 credits (33,578,023)PSP LLR Turquoise: Earned 5,000,000 credits (6,587,988)SoB LLR Sapphire: Earned 20,000,000 credits (44,987,733)SR5 LLR Turquoise: Earned 5,000,000 credits (6,205,694)SGS LLR Ruby: Earned 2,000,000 credits (3,627,819)TRP LLR Turquoise: Earned 5,000,000 credits (7,078,152)Woodall LLR Amethyst: Earned 1,000,000 credits (1,693,614)321 Sieve (suspended) Emerald: Earned 50,000,000 credits (50,256,050)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,571,178)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,009,610)PPS Sieve Double Silver: Earned 200,000,000 credits (455,672,175)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,165,888)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,071,454)AP 26/27 Turquoise: Earned 5,000,000 credits (6,798,063)GFN Emerald: Earned 50,000,000 credits (57,113,430)WW Ruby: Earned 2,000,000 credits (2,072,000)PSA Double Bronze: Earned 100,000,000 credits (102,762,384)
Message 155480 - Posted: 19 May 2022 | 2:29:47 UTC - in response to Message 155476.

How on earth did you persuade Boinc to use more threads than you have?

You could fool BOINC into thinking that the app uses less than a full core by setting up app_config.xml
There are plenty of postings around here on how to do that.
I haven't done that lately, and I only used it to run multiple GPU tasks simultaneously.

Is that by using something like I did here?

<app_version> <app_name>milkyway_nbody</app_name> <plan_class>mt</plan_class> <avg_ncpus>4</avg_ncpus> <cmdline>--nthreads 12</cmdline> </app_version>

The app can get 12 cores, but since it uses an average of 4, I tell boinc this, so it runs more of them.


Yes, exactly. Do an advanced search in the forums for app_config.xml to find settings for PrimeGrid apps.

Peter Hucker of the Scottish Boinc Team
Send message
Joined: 23 Sep 06
Posts: 144
ID: 3541
Credit: 68,069,390
RAC: 14,398
321 LLR Silver: Earned 100,000 credits (117,162)Cullen LLR Silver: Earned 100,000 credits (199,800)ESP LLR Silver: Earned 100,000 credits (100,301)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (100,048)PPS LLR Bronze: Earned 10,000 credits (31,226)PSP LLR Silver: Earned 100,000 credits (296,161)SoB LLR Amethyst: Earned 1,000,000 credits (1,468,570)SR5 LLR Bronze: Earned 10,000 credits (44,401)SGS LLR Silver: Earned 100,000 credits (107,261)TRP LLR Bronze: Earned 10,000 credits (62,426)Woodall LLR Silver: Earned 100,000 credits (240,222)321 Sieve (suspended) Bronze: Earned 10,000 credits (30,838)PPS Sieve Turquoise: Earned 5,000,000 credits (5,004,308)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Bronze: Earned 10,000 credits (16,965)TRP Sieve (suspended) Bronze: Earned 10,000 credits (14,492)AP 26/27 Jade: Earned 10,000,000 credits (12,275,843)GFN Sapphire: Earned 20,000,000 credits (45,717,763)WW Ruby: Earned 2,000,000 credits (2,200,000)
Message 155481 - Posted: 19 May 2022 | 5:51:18 UTC

I do wish this page had more details, it mentions nthreads but doesn't say what it does!

https://boinc.berkeley.edu/wiki/Client_configuration

Their instructions, just like their programming, leaves a lot to be desired.

Post to thread

Message boards : Number crunching : Multithreading?

[Return to PrimeGrid main page]
DNS Powered by DNSEXIT.COM
Copyright © 2005 - 2022 Rytis Slatkevičius (contact) and PrimeGrid community. Server load 0.99, 1.19, 0.93
Generated 26 Jun 2022 | 5:11:57 UTC