mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   GPU to 72 (https://www.mersenneforum.org/forumdisplay.php?f=95)
-   -   GPU to 72 status... (https://www.mersenneforum.org/showthread.php?t=16263)

dragonbud20 2015-09-10 23:44

[QUOTE=airsquirrels;410019]It must be calamity time.[/QUOTE]
Looks like somethings up... SoCal's in a heatwave so I can't run above about 50% capacity without cooking either myself or my computer

chalsall 2015-09-10 23:50

[QUOTE=dragonbud20;410055]Looks like somethings up... SoCal's in a heatwave so I can't run above about 50% capacity without cooking either myself or my computer[/QUOTE]

LOL... Don't cook either. :smile:

So everyone knows, LG72D is currently sometimes assigning TF'ing work to 74 to hand out to P-1'ers.

Not optimal, but better than at 73.

Madpoo 2015-09-11 04:04

[QUOTE=dragonbud20;410055]Looks like somethings up... SoCal's in a heatwave so I can't run above about 50% capacity without cooking either myself or my computer[/QUOTE]

I'm in LA myself right now, installing some stuff into a datacenter.

Now, I've been in colocations around the country and have some experience in the cold/warm aisles. I have to say though that the [B]humidity[/B] in this LA location is frustrating!

I like to wrap my cables with id labels that have a number on them, so I can easily trace what goes where. Well, the darn labels get all gummy from the high humidity and peel off after a few hours. It's crazy. I don't know how people here can stand it. :smile:

That said, at least the ambient temps are kept to 70'ish F. But that RH must be pretty high... seems like > 50% in there, maybe > 60% outside? Good gravy.

I know... the tropically located folks in here are going to one-up those figures with their tales of 99% humidity, and now I can feel your pain a little bit more. :smile:

frmky 2015-09-11 06:09

It's not normally THIS humid here. But this week and especially today have been miserable.

Mark Rose 2015-09-11 13:05

[QUOTE=chalsall;410056]So everyone knows, LG72D is currently sometimes assigning TF'ing work to 74 to hand out to P-1'ers.

Not optimal, but better than at 73.[/QUOTE]

What's the best way to tell where the P-1 wave front is? I think it would be super useful if there were a page that showed where the LL and P-1 wave fronts are according to GPU72, including perhaps the buffer.

I'm also curious how far behind LL is from P-1, then we could figure out the amount of necessary work to catch up to the P-1 wave, which would allow going to 75 before P-1'ing.

James Heinrich 2015-09-11 14:05

[QUOTE=Madpoo;410073]at least the ambient temps are kept to 70'ish F. But that RH must be pretty high... seems like > 50% in there, maybe > 60% outside? Good gravy.[/QUOTE]Sounds like a normal summer 1000+km north in Ontario, ~27x°C (80F) with 75% humidity feels like 36°C (97F) is a perfectly normal August day. And six months later it's 70°C (160F) colder :rajula:

dragonbud20 2015-09-11 14:14

Yeah the humidity this week has been pretty bad. Normally it's around 20% so this is rather high especially for this time of year.

chalsall 2015-09-11 14:25

[QUOTE=James Heinrich;410092]Sounds like a normal summer 1000+km north in Ontario, ~27x°C (80F) with 75% humidity feels like 36°C (97F) is a perfectly normal August day.[/QUOTE]

LOL... The other day here it was 31 degrees with 86% humidity... "Felt like" 43 degrees! My workmen and I were melting!!!

chalsall 2015-09-11 14:40

[QUOTE=Mark Rose;410090]What's the best way to tell where the P-1 wave front is? I think it would be super useful if there were a page that showed where the LL and P-1 wave fronts are according to GPU72, including perhaps the buffer.[/QUOTE]

Probably the best report to look at would be the [URL="https://www.gpu72.com/reports/available/"]Available Assignments[/URL] page; you can see that the P-1 "wave back" is at 72M. If you click on the "No P-1" column header you can see we don't have much at 75 nor 74. Keep in mind that GPU72 releases some candidates at 75 (or 74 if we're "starved") without P-1 done for Primenet to assign to P-1'ers.

[QUOTE=Mark Rose;410090]I'm also curious how far behind LL is from P-1, then we could figure out the amount of necessary work to catch up to the P-1 wave, which would allow going to 75 before P-1'ing.[/QUOTE]

For that, take a look at the [URL="https://www.gpu72.com/reports/estimated_completion/primenet/"]Estimated Completion[/URL] report. The P-1'ers are about 430 days ahead of the LL'ers.

Mark Rose 2015-09-11 15:11

[QUOTE=chalsall;410095]If you click on the "No P-1" column header you can see we don't have much at 75 nor 74. Keep in mind that GPU72 releases some candidates at 75 (or 74 if we're "starved") without P-1 done for Primenet to assign to P-1'ers.[/quote]

I never thought to click on those. I always thought they were just sorting headers, not to new pages :)

[quote]For that, take a look at the [URL="https://www.gpu72.com/reports/estimated_completion/primenet/"]Estimated Completion[/URL] report. The P-1'ers are about 430 days ahead of the LL'ers.[/QUOTE]

Thanks :)

dragonbud20 2015-09-12 19:37

[QUOTE=chalsall;410095]Keep in mind that GPU72 releases some candidates at 75 (or 74 if we're "starved") without P-1 done for Primenet to assign to P-1'ers.[/QUOTE]

Why is it done like this? I assume it's because from a computing power/time required standpoint it easier to trial factor below a certain point and easier to P-1 past that and at the current exponent the bit depth line is around 75? did I get any of that right?


All times are UTC. The time now is 23:16.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.