[QUOTE=chalsall;345375] it was a stupid human error.[/QUOTE]

You mean a SHE, vs a SPE. Is that supposed to mean programmers aren´t humans? :wink:

[QUOTE=lycorn;345379]You mean a SHE, vs a SPE. Is that supposed to mean programmers aren´t humans? :wink:[/QUOTE]

To most women, programmers aren't quite human. Thank goodness for those women who are a little more tolerant.... :smile:

We can pick up LL tests from GPU72.com?

EDIT: Derp. CudaLucas. Derp dee derp dee derp.

Dap, you can do that automatically too, with P95 I mean, by setting the right proxy in Test/PrimeNet/Connection menu. That should be something like gimps.gpu72.com, port 80 (if I am not mistaken; someone else please confirm). The advantage is that gpu72's spiders check PrimeNet server periodically and keep only smallest assignments to offer (smallest expo goes to the most trusted user, i.e. the one who returns the results in time).

Someone just went [URL="https://www.gpu72.com/reports/workers/"]over[/URL] the MILLION GHz days...

[QUOTE=kracker;346773]Someone just went [URL="https://www.gpu72.com/reports/workers/"]over[/URL] the MILLION GHz days...[/QUOTE]

Congratulations to old What's-His-Name! :cmd:

[QUOTE=kracker;346773]Someone just went [URL="https://www.gpu72.com/reports/workers/"]over[/URL] the MILLION GHz days...[/QUOTE]

I was thinking about this too — congratulations Jerry.

Thanks ;-) Couldn't do it without a lot of help from everyone who has helped setup G72 and also helped automate everything, so thanks to everyone else as well.

impressive
 

Nice work Flash!

Jerry: What machinery are you running?

Cool.
 

Thanks for everything Jerry!

[QUOTE=TheMawn;346823]Jerry: What machinery are you running?[/QUOTE]
Currently I have:

7 - MSI GTX 580 Lightning running MFAKTC
2 - i7 3770k systems running P-1 on P95 (one with 32Gb, one with 16Gb)
1 - Dual 6272 Opteron System running P-1 on P95 with 128Gb ECC
and a few misc systems running this or that

The system with 32Gb is a Quad MSI GTX 580 Lightning all in a watercooled loop, but one of the cards is broken so it sits idle for now until I get a chance to swap it out someday.

[QUOTE=flashjh;346840]Currently I have:

7 - MSI GTX 580 Lightning running MFAKTC
2 - i7 3770k systems running P-1 on P95 (one with 32Gb, one with 16Gb)
1 - Dual 6272 Opteron System running P-1 on P95 with 128Gb ECC
and a few misc systems running this or that

The system with 32Gb is a Quad MSI GTX 580 Lightning all in a watercooled loop, but one of the cards is broken so it sits idle for now until I get a chance to swap it out someday.[/QUOTE]
How much do you pay the electric company for [B][I]7[/I][/B] 580's??

[SIZE=1]There should be a thread on people showing/saying their hardware for crunching, I would be very interested in seeing what others have... :razz: Not for me of course, my hardware sucks and is really low end...[/SIZE]

[QUOTE=flashjh;346840]Currently I have:

7 - MSI GTX 580 Lightning running MFAKTC
2 - i7 3770k systems running P-1 on P95 (one with 32Gb, one with 16Gb)
1 - Dual 6272 Opteron System running P-1 on P95 with 128Gb ECC
and a few misc systems running this or that

The system with 32Gb is a Quad MSI GTX 580 Lightning all in a watercooled loop, but one of the cards is broken so it sits idle for now until I get a chance to swap it out someday.[/QUOTE]

I would have guessed from your avatar that you have a basement full of Commodore 64's :toot:

Server's down
 

Seems that the server is down....
[url]http://www.downforeveryoneorjustme.com/http://www.gpu72.com/[/url]

[QUOTE=Uncwilly;347028]Seems that the server is down....
[url]http://www.downforeveryoneorjustme.com/http://www.gpu72.com/[/url][/QUOTE]

It's just you...

Uptime shows 160 days.

[url]http://gpu72.com[/url] works for me too. It seems the site checker isn't 100% reliable as it often mistakenly thinks Google is down.

[QUOTE=chalsall;347040]It's just you...

Uptime shows 160 days.[/QUOTE]

@UNC- you might add isup.me to your lineup of "on/off check" sites.

[QUOTE=Uncwilly;347028]Seems that the server is down....[/QUOTE]
[QUOTE=chalsall;347040]It's just you...

Uptime shows 160 days.[/QUOTE]
I tried several times over the course of an hour. Then I checked dfeojm it agreed. Next time I checked, about 1/4-1/2 hour after posting it was ok.

[QUOTE=Uncwilly;347044]I tried several times over the course of an hour. Then I checked dfeojm it agreed. Next time I checked, about 1/4-1/2 hour after posting it was ok.[/QUOTE]

OK... All I can tell you is this:

[CODE][22/Jul/2013:18:23:12 -0400] www.gpu72.com 50.97.161.229 - - "HEAD / HTTP/1.1" 200 - "-" "SiteCheck - http://downforeveryoneorjustme.com"
[22/Jul/2013:18:23:13 -0400] www.gpu72.com 50.97.161.229 - - "HEAD / HTTP/1.1" 200 - "-" "SiteCheck - http://downforeveryoneorjustme.com"
[22/Jul/2013:18:23:13 -0400] www.gpu72.com 50.97.161.229 - - "HEAD / HTTP/1.1" 200 - "-" "SiteCheck - http://downforeveryoneorjustme.com"
[22/Jul/2013:18:30:24 -0400] www.gpu72.com 50.97.161.229 - - "HEAD / HTTP/1.1" 200 - "-" "SiteCheck - http://downforeveryoneorjustme.com"
[22/Jul/2013:18:30:26 -0400] www.gpu72.com 50.97.161.229 - - "HEAD / HTTP/1.1" 200 - "-" "SiteCheck - http://downforeveryoneorjustme.com"
[22/Jul/2013:18:32:57 -0400] www.gpu72.com 50.97.161.229 - - "HEAD / HTTP/1.1" 200 - "-" "SiteCheck - http://downforeveryoneorjustme.com"
[22/Jul/2013:18:32:57 -0400] www.gpu72.com 50.97.161.229 - - "HEAD / HTTP/1.1" 200 - "-" "SiteCheck - http://downforeveryoneorjustme.com"
[22/Jul/2013:18:36:36 -0400] www.gpu72.com 50.97.161.229 - - "HEAD / HTTP/1.1" 200 - "-" "SiteCheck - http://downforeveryoneorjustme.com"
[22/Jul/2013:18:36:37 -0400] www.gpu72.com 50.97.161.229 - - "HEAD / HTTP/1.1" 200 - "-" "SiteCheck - http://downforeveryoneorjustme.com"
[22/Jul/2013:18:45:45 -0400] gpu72.com 50.97.161.229 - - "HEAD / HTTP/1.1" 301 - "-" "SiteCheck - http://downforeveryoneorjustme.com"
[22/Jul/2013:18:58:32 -0400] www.gpu72.com 50.97.161.229 - - "HEAD / HTTP/1.1" 200 - "-" "SiteCheck - http://downforeveryoneorjustme.com"
[22/Jul/2013:18:58:32 -0400] www.gpu72.com 50.97.161.229 - - "HEAD / HTTP/1.1" 200 - "-" "SiteCheck - http://downforeveryoneorjustme.com"
[/CODE]

[QUOTE=Uncwilly;347044]I tried several times over the course of an hour. Then I checked dfeojm it agreed. Next time I checked, about 1/4-1/2 hour after posting it was ok.[/QUOTE]

I have to second Unc's opinion, though I just waited and did not seek a second opinion. Perhaps he and I are on the same plugged tube in the interwebs? Both View Assignments and Individual Overall Statistics were giving me this:

[QUOTE][B]Internal Server Error[/B]

The server encountered an internal error or misconfiguration and was unable to complete your request.
Please contact the server administrator, [EMAIL="chalsall@ideas4lease.com"]chalsall@ideas4lease.com[/EMAIL] and inform them of the time the error occurred, and anything you might have done that may have caused the error.
More information about this error may be available in the server error log.
Apache/2.2.3 (CentOS) Server at [URL="http://www.gpu72.com"]www.gpu72.com[/URL] Port 80[/QUOTE]

I got that same message a while ago as well, seems to be fine now.

Straight after he posted I checked his link and it said down.

[QUOTE=henryzz;347108]Straight after he posted I checked his link and it said down.[/QUOTE]

Sucks to be me, huh?

[QUOTE=chalsall;347133]Sucks to be me, huh?[/QUOTE]

Actually, doesn't it suck to have users with bad DNS, behind plugged up routers? I am quite happy to go with the "clogged tubes in the interwebs" theory.

Is there a reason that the 62M LL range at 73 bit never gets assigned? :max:

If one goes to the request page and leave everything with the default values, we get exponent in the 63M range to factor from 73 to 74. We need to specify the optional range min and max value to get exponent in the 62M range. :cry:

I think we should release 62M at 73 bits. People are being assigned LL tests in the 66M range. Clearly we are having trouble keeping the pace. A bit of breathing room is required.

[QUOTE=garo;352229]I think we should release 62M at 73 bits. People are being assigned LL tests in the 66M range. Clearly we are having trouble keeping the pace. A bit of breathing room is required.[/QUOTE]

Is that an order Commander? My instruments show that things are cool at 74, despite some stray impacts...

Can we wait for Captain Jean-Luc Picard (read: George) for the final authoritative order? The firing solution is programmed... (That's meant to be serious, and funny, at the same time....)

[QUOTE=garo;352229]I think we should release 62M at 73 bits. People are being assigned LL tests in the 66M range. Clearly we are having trouble keeping the pace. A bit of breathing room is required.[/QUOTE]

What is easy to determine is a good percentage of these assignments belong to Mr. A.Nony with no reported progress. That is a fact. Any proposed interpretation would only be a theory.
What I can not easily determine is whether most or all assigned for LL as deep as 66M are already factored to 73 or 74 bits. I leave that the the better informed.

[QUOTE=chalsall;352231]Is that an order Commander? My instruments show that things are cool at 74, despite some stray impacts...

Can we wait for Captain Jean-Luc Picard (read: George) for the final authoritative order? The firing solution is programmed... (That's meant to be serious, and funny, at the same time....)[/QUOTE]

Then why are we getting 66M exponents for testing? Where is the actual "wave" at now?(removing any that gpu72 has for tf'ing)
Just a random thought :smile:

[QUOTE=petrw1;352232]What is easy to determine is a good percentage of these assignments belong to Mr. A.Nony with no reported progress.[/QUOTE]Anyone check his IP :davieddy:

[QUOTE=diamonddave;352216]Is there a reason that the 62M LL range at 73 bit never gets assigned?[/QUOTE]

End of the week. Answering questions here...

The 62M range does get assigned for TFing. However, those candidates without P-1 have preference because we've suddenly found ourselves with so much P-1 ability. Surfing the waves...

[QUOTE=diamonddave;352216]If one goes to the request page and leave everything with the default values, we get exponent in the 63M range to factor from 73 to 74. We need to specify the optional range min and max value to get exponent in the 62M range. :cry:[/QUOTE]

Or, on the other hand, you can simply click on the drop down menu and select "Lowest Candidate" (or the MISFIT equivalent)... This option has been available for quite some time, but only recently did it also mean "possibly already had P-1 done"....

[QUOTE=Uncwilly;352234]Anyone check his IP :davieddy:[/QUOTE]

No, don't go there. This is not him.

This is nominal.

[QUOTE=kracker;352233]Then why are we getting 66M exponents for testing? Where is the actual "wave" at now?[/QUOTE]

You have read Heisenberg? (:smile:)

[QUOTE=chalsall;352237]You have read Heisenberg? (:smile:)[/QUOTE]
No NO! Don't look in the box! You'll kill the CAT!:max:

[QUOTE]Or, on the other hand, you can simply click on the drop down menu and select "Lowest Candidate" (or the MISFIT equivalent)...[/QUOTE]

That has been one of my choices for a long time. If I'm going to do last bit factoring, I'd rather it was the lowest exponents getting cleared. On the other hand, sometimes I like to take things from less than a 73 bit starting point so that I can hope to see an occasional factor. I look at the available assignments and request accordingly.

Then there's the 332.xM range..... I see that as helping out people who want to take on crazy-large LLs.

The only place I haven't gone lately is DCTF.

I really don't understand P-1. Does having more memory speed up the process or increase the range of prime factors that might be found? Having 1,000 GHz-Days of P-1 puts you in the top 100. Having 10,000 GHz-Days puts you in the top 10. Obviously we don't need very much P-1 if P-1 is keeping ahead of TF.

[QUOTE=chalsall;352231]Can we wait for Captain Jean-Luc Picard (read: George) for the final authoritative order? The firing solution is programmed... (That's meant to be serious, and funny, at the same time....)[/QUOTE]

I do not have a strong opinion. If we are currently assigning 66M LLs that are only TFed to 73, then we should be assigning the 63Ms instead. If the 66M assignments are TFed to 2^74 and Chris believes he has the firepower to both keep ahead of the LL wavefront and get the 63Ms to 2^74, then I'm OK with holding the 63Ms back.

Shouldn't factoring 63M to 74 be the priority then? Or is that the thing about some user having a whole bunch of 63M reserved and not handing in results?

[QUOTE=TheMawn;352244]I really don't understand P-1. Does having more memory speed up the process or increase the range of prime factors that might be found? Having 1,000 GHz-Days of P-1 puts you in the top 100. Having 10,000 GHz-Days puts you in the top 10. Obviously we don't need very much P-1 if P-1 is keeping ahead of TF.[/QUOTE]
Um, compare GHz days on doing a exponent with P-1 vs TF?

My P-1s are all ~63.5M. They yield ~ 4 GHz-days apiece.

A ~63M TF from 73-74 coughs up ~30 GHz-days. I crank out a lot more of those. Such are the joys of 2 GPUs. The P-1s are something to keep the CPU cores busy in the meantime.

[QUOTE=Prime95;352245]I do not have a strong opinion. If we are currently assigning 66M LLs that are only TFed to 73, then we should be assigning the 63Ms instead. If the 66M assignments are TFed to 2^74 and Chris believes he has the firepower to both keep ahead of the LL wavefront and get the 63Ms to 2^74, then I'm OK with holding the 63Ms back.[/QUOTE]

Copy. Thank you.

And trust us....

[QUOTE=chalsall;352254]Copy. Thank you.

And trust us....[/QUOTE]

Besides... If he messes up, you can just turn Makes Sense[SUP]TM[/SUP] assign double-checks until we catch up :razz:

[QUOTE=TheMawn;352244]I really don't understand P-1. Does having more memory speed up the process or increase the range of prime factors that might be found?[/QUOTE]

If you ask that questions for fixed bounds (B1 and B2), then more memory speeds up the stage 2. But Prime95 takes the amount of memory allowed into account when it chooses the bounds. So in short, the answer is "maybe both, maybe just one, and maybe neither".

Having more memory speeds up stage 2. If you have more memory, then, you increase the range of prime factors that might be found, while the time required may go up or down (whatever it takes to find factors most efficiently).

If you have so little memory allowed that Prime95 decides to skip stage 2, it will choose a slightly higher B1 (i.e. a longer stage 1), but the overall test will take less time than with a stage 2.

So in practice, more memory might actually make the P-1 for a single number take longer, but with a higher chance of finding a factor, which means better efficiency for you and GIMPS (more factors cleared per time spent).

(something interesting I just realized: there's a slim chance that for a given exponent, you'd find a factor by assigning the minimum memory and skipping stage 2, and you'd miss that factor if you assign a generous amount of memory and thus choose a lower B1; this would occur if, e.g. the former B1=B2=660 000 and the latter B1=570 000, B2=11 000 000 and the k in the factor has 2 or more factors n with 570 000 < n < 660 000)

[QUOTE=TheMawn;352244]I really don't understand P-1. Does having more memory speed up the process or increase the range of prime factors that might be found?[/QUOTE]It can be either, depending on circumstances.

[URL="http://www.mersenneforum.org/showpost.php?p=153984&postcount=71"]Here's[/URL] one explanation of the two different ways of invoking P-1 and their different relationships between B1/B2 bounds, "available memory", speed, and probability of finding a factor.

[URL="http://www.mersenneforum.org/showpost.php?p=209603&postcount=12"]Here's[/URL] another version.

[I still haven't posted a grand amalgamated version at mersennewiki.org :-( ]

Looks like Primenet is assigning 67Ms. SPE chalsall?

[QUOTE=garo;352830]Looks like Primenet is assigning 67Ms. SPE chalsall?[/QUOTE]

Nope.

Instead, very careful management. Note all those assignments are TFed to at least 74, and P-1'ed.

[QUOTE=garo;352830]Looks like Primenet is assigning 67Ms. SPE chalsall?[/QUOTE]

SPE? (I'm sure it's obvious, but the meaning eludes me.)

[URL]http://en.wikipedia.org/wiki/SPE[/URL]
[QUOTE][B]Science and technology[/B]

[LIST][*][URL="http://en.wikipedia.org/wiki/Serum_protein_electrophoresis"]Serum protein electrophoresis[/URL][*][URL="http://en.wikipedia.org/wiki/Solar_proton_event"]Solar proton event[/URL], in which protons emitted by the Sun become accelerated to very high energies[*][URL="http://en.wikipedia.org/wiki/Solid_phase_extraction"]Solid phase extraction[/URL], a separation process used to isolate compounds from a mixture[*][URL="http://en.wikipedia.org/wiki/Epitaxy#Solid-phase"]Solid-phase epitaxy[/URL], a method for forming crystalline materials from amorphous precursors[*][I][URL="http://en.wikipedia.org/wiki/The_Sound_Pattern_of_English"]The Sound Pattern of English[/URL][/I], in phonology, a book by Noam Chomsky and Morris Halle, and method of notation of phonological rules[*][URL="http://en.wikipedia.org/wiki/Stanford_prison_experiment"]Stanford prison experiment[/URL], in psychology, an infamous study of prisoner and prison guard roles[*][URL="http://en.wikipedia.org/wiki/Subgame_perfect_equilibrium"]Subgame perfect equilibrium[/URL], in game theory, a refinement of a Nash equilibrium used in dynamic games[*]Synchronous payload envelope, in telecommunications, related to [URL="http://en.wikipedia.org/wiki/Synchronous_optical_networking"]synchronous optical networking[/URL][/LIST] [B]Computer science[/B]

[LIST][*][URL="http://en.wikipedia.org/wiki/Stani%27s_Python_Editor"]Stani's Python Editor[/URL], a multi-platform IDE for the Python programming language[*]Synergistic Processing Element, part of the [URL="http://en.wikipedia.org/wiki/Cell_%28microprocessor%29#Synergistic_Processing_Elements_.28SPE.29"]Cell[/URL] microprocessor used in the PlayStation 3 video game console[*]Software performance engineering, in [URL="http://en.wikipedia.org/wiki/Performance_engineering"]performance engineering[/URL][*]Signal-Processing Engine, for example in the [URL="http://en.wikipedia.org/wiki/PowerPC_e500"]PowerPC e500[/URL][/LIST] [B]Organizations[/B]

[LIST][*][URL="http://en.wikipedia.org/wiki/Schlumberger_Limited"]Schlumberger Limited[/URL], formerly Société de Prospection Electrique, the world's largest oilfield services corporation[*][URL="http://en.wikipedia.org/wiki/Sigma_Phi_Epsilon"]Sigma Phi Epsilon[/URL], a fraternity for male college students in the United States[*][URL="http://en.wikipedia.org/wiki/Societas_Privata_Europaea"]Societas Privata Europaea[/URL], a Europe-wide form of a limited company[*][URL="http://en.wikipedia.org/wiki/Sony_Pictures_Entertainment"]Sony Pictures Entertainment[/URL], an American-based entertainment company owned by media corporation Sony[*][URL="http://en.wikipedia.org/wiki/SPE_%28company%29"]SPE (company)[/URL], former name of the Belgian energy company EDF Luminus[*][URL="http://en.wikipedia.org/wiki/Special_Police_Establishment"]Special Police Establishment[/URL], predecessor of Central Bureau of Investigation, India[*][URL="http://en.wikipedia.org/wiki/Special_purpose_entity"]Special purpose entity[/URL], a legal entity created to fill a narrow or temporary objective[/LIST] [B]Professional societies[/B]

[LIST][*][URL="http://en.wikipedia.org/wiki/Society_for_Photographic_Education"]Society for Photographic Education[/URL][*][URL="http://en.wikipedia.org/wiki/Society_of_Petroleum_Engineers"]Society of Petroleum Engineers[/URL][LIST][*][URL="http://en.wikipedia.org/wiki/SPE_Certified_Petroleum_Professional"]SPE Certified Petroleum Professional[/URL], a designation[/LIST] [*][URL="http://en.wikipedia.org/wiki/Society_of_Plastics_Engineers"]Society of Plastics Engineers[/URL][*][URL="http://en.wikipedia.org/wiki/Society_of_Professional_Engineers_UK"]Society of Professional Engineers UK[/URL][/LIST] [B]Other[/B]

[LIST][*]Supervisory patent examiner, a [URL="http://en.wikipedia.org/wiki/Patent_examiner"]patent examiner[/URL], usually a civil servant, working at a patent office[*][URL="http://en.wikipedia.org/wiki/SPE_Certified"]SPE Certified[/URL], a food quality assurance certification[/LIST][/QUOTE]

In this context SPE is Stupid Programmer Error ([SUP]TM[/SUP] chalsall).

[QUOTE=sdbardwick;352838]In this context SPE is Stupid Programmer Error ([SUP]TM[/SUP] chalsall).[/QUOTE]

Of Course! DUH! :blush:

Synonym:
[quote=wikipedia][B]Computer science[/B]

[LIST][*]Software performance engineering ...[URL="http://en.wikipedia.org/wiki/Performance_engineering"]
[/URL][/LIST][/quote]

100,000 P1. 1,000,000 GHz Days saved
 

Not me....G72.

1,000,000 GHz-Days of LL tests alone is a LOT. That's almost twice as much as gets done in an entire year. About four times Curtisc's yearly throughput.

Go team GPU!

[QUOTE=TheMawn;353207]1,000,000 GHz-Days of LL tests alone is a LOT. That's almost twice as much as gets done in an entire year. About four times Curtisc's yearly throughput.

Go team GPU![/QUOTE]

Actually, four times curtisc's yearly throughput would be [B]10[/B],000,000 GHz-days. But what's an order of magnitude amongst friends? :smile:

One million GHz-days, or 1 Petahertz-day, or just under 2740 GHz-years, however, is certainly still an accomplishment of which to be quite proud. Huzzah!

On 2013-01-22, I went through and added up all of the GHz-days totals in the account rankings for the last year, and then over every year back to 1996. I was quite shocked to find that some 28.5% of GIMPS all-time throughput had been achieved in the most recent 12 months! (For the number crunchers among us, I calculated 58,509,783 GHz-days since the Dawn-of-GIMPS(TM), and 16,678,308 GHz-days in the period 2012-01-22 to 2013-01-22.)

1 GHz-day = 5.075 P-90 years (original GIMPS work measurement). Assuming a single stock GTX 580 processes [url=http://www.mersenne.ca/mfaktc.php]433.2 GHz-days/day[/url] (running mfaktc), that single GPU will do what took George's original Pentium 90 an entire [i]year[/i] of work... every 17 minutes.

[QUOTE=James Heinrich;353230]1 GHz-day = 5.075 P-90 years (original GIMPS work measurement). Assuming a single stock GTX 580 processes [url=http://www.mersenne.ca/mfaktc.php]433.2 GHz-days/day[/url] (running mfaktc), that single GPU will do what took George's original Pentium 90 an entire [i]year[/i] of work... every 17 minutes.[/QUOTE]

How things change; how they stay the same.

[QUOTE=James Heinrich;353230]1 GHz-day = 5.075 P-90 years (original GIMPS work measurement). Assuming a single stock GTX 580 processes [URL="http://www.mersenne.ca/mfaktc.php"]433.2 GHz-days/day[/URL] (running mfaktc), that single GPU will do what took George's original Pentium 90 an entire [I]year[/I] of work... every 17 minutes.[/QUOTE]

Are you sure you have this right? Check your calculation again.

[QUOTE=garo;353258]Are you sure you have this right? Check your calculation again.[/QUOTE]

I get [SPOILER]39.3 seconds[/SPOILER].

[QUOTE=garo;353258]Are you sure you have this right? Check your calculation again.[/QUOTE]

Commander...

What, please, exactly, is your point?

Whoops I saw 250k GHz-Days, not 2.5M. Haha. D'oh well. Still a great accomplishment.

[QUOTE=James Heinrich;353230]1 GHz-day = 5.075 P-90 years (original GIMPS work measurement). Assuming a single stock GTX 580 processes [URL="http://www.mersenne.ca/mfaktc.php"]433.2 GHz-days/day[/URL] (running mfaktc), that single GPU will do what took George's original Pentium 90 an entire [I]year[/I] of work... every 17 minutes.[/QUOTE]

Other way around: 1 P-90 year = 5.075 GHz-days.

So 433.2 GHz-days/day = 85.36 P-90 years/day = 1 P-90 year of work completed every 1/85.36 * 1440 = 16.87 minutes.

So assume the same level of progress in computers, 20 years from now, how long would it take to do all the work G72 has done?

GPU72 has a yearly rate of 2,500,000 GHz-Days as of now. Let's forget the 1,000,000 GHz-Days of saved work. We could factor it in later quite easily.

The thing to keep in mind is we're looking at GPU vs CPU which isn't exactly fair as the GPU architecture is extremely well suited to some of the computing we're doing. GPU's aren't exactly an improvement on a CPU.

Considering the improvements CPU-wise, where a 4.6 GHz Core i5-3570k is worth 12 GHz of Core 2, which in turn is 2.4 P90-years, or 876 P90-days. A single core of mine is 876 times faster than the P90. 3504 times faster if we count the increase in core count, but let us not.

If the same level of increase can be attributed to a GTX 580, then a 876-fold increase over 430 GHz-Days per day gives 376,780 GHz-Days per day. Divide the yearly GPU72 production by this value and you get 6.635.

In other words, a high end (compute-wise the GTX 580 is probably close to a GTX 780) GPU in twenty years will likely have the capacity to beat the entire GPU72 team's yearly output in one week.

[QUOTE=TheMawn;353382]GPU72 has a yearly rate of 2,500,000 GHz-Days as of now. Let's forget the 1,000,000 GHz-Days of saved work. We could factor it in later quite easily.

The thing to keep in mind is we're looking at GPU vs CPU which isn't exactly fair as the GPU architecture is extremely well suited to some of the computing we're doing. GPU's aren't exactly an improvement on a CPU.

Considering the improvements CPU-wise, where a 4.6 GHz Core i5-3570k is worth 12 GHz of Core 2, which in turn is 2.4 P90-years, or 876 P90-days. A single core of mine is 876 times faster than the P90. 3504 times faster if we count the increase in core count, but let us not.

If the same level of increase can be attributed to a GTX 580, then a 876-fold increase over 430 GHz-Days per day gives 376,780 GHz-Days per day. Divide the yearly GPU72 production by this value and you get 6.635.

In other words, a high end (compute-wise the GTX 580 is probably close to a GTX 780) GPU in twenty years will likely have the capacity to beat the entire GPU72 team's yearly output in one week.[/QUOTE]

?

I didn't know there was a yearly done and saved on gpu72... I just see lifetime.

Calculations for the gpu72 [I]team [/I]on Primenet seem right, but many aren't in the team.

I wonder if Curtis Cooper knows about this sub-project. Surely there must be GPU-enabled workstations at UCM?

[QUOTE=ixfd64;353389]I wonder if Curtis Cooper knows about this sub-project. Surely there must be GPU-enabled workstations at UCM?[/QUOTE]

Running mfaktx at higher bit levels while computer is in use means lag, though...

[QUOTE=TheMawn;353382]GPU72 has a yearly rate of 2,500,000 GHz-Days as of now. Let's forget the 1,000,000 GHz-Days of saved work. We could factor it in later quite easily.

The thing to keep in mind is we're looking at GPU vs CPU which isn't exactly fair as the GPU architecture is extremely well suited to some of the computing we're doing. GPU's aren't exactly an improvement on a CPU.

Considering the improvements CPU-wise, where a 4.6 GHz Core i5-3570k is worth 12 GHz of Core 2, which in turn is 2.4 P90-years, or 876 P90-days. A single core of mine is 876 times faster than the P90. 3504 times faster if we count the increase in core count, but let us not.

If the same level of increase can be attributed to a GTX 580, then a 876-fold increase over 430 GHz-Days per day gives 376,780 GHz-Days per day. Divide the yearly GPU72 production by this value and you get 6.635.

In other words, a high end (compute-wise the GTX 580 is probably close to a GTX 780) GPU in twenty years will likely have the capacity to beat the entire GPU72 team's yearly output in one week.[/QUOTE]
That is PROBABLY not going to happen, transistors can't get that small. We are now at 28nm, which is about 100-150 atoms. For GPUs to get 876 times faster, you need either 876x higher clock speeds (no way that is going to happen). Or 876x amount of transistors, and in order to pack those transistors on a GPU, they need to be very small, and we will have huge difficulties past 10nm (due to quantum tunnelling).

[QUOTE=VictordeHolland;353472]That is not going to happen....[/QUOTE]

How can you say that with such certainty?

[QUOTE=chalsall;353475]How can you say that with such certainty?[/QUOTE]
Bought a timemachine from Ebay, travelled to the year 2033, asked the robots (who wiped out mankind): how many transistors are there in your body, asked them for the next Mersenne Primes (which they wouldn't give to me), spoke to a half a dozen aliens, travelled back into time, woke up with a headache.

[QUOTE=VictordeHolland;353482]Bought a timemachine from Ebay, travelled to the year 2033, asked the robots (who wiped out mankind): how many transistors are there in your body, asked them for the next Mersenne Primes (which they wouldn't give to me), spoke to a half a dozen aliens, travelled back into time, woke up with a headache.[/QUOTE]

Actually, compared to some of the answers I've been receiving recently, this is actually passable... 8-)

[QUOTE=chalsall;353492]Actually, compared to some of the answers I've been receiving recently, this is actually passable... 8-)[/QUOTE]
+1

[QUOTE=chalsall;353492]Actually, compared to some of the answers I've been receiving recently, this is actually passable... 8-)[/QUOTE]

LOLOL...
+1

[QUOTE=chalsall;353492]Actually, compared to some of the answers I've been receiving recently, this is actually passable... 8-)[/QUOTE]

Actually? :devil:
EDIT: But seriously, I agree that it's an amusingly inventive answer.

EDIT2: [QUOTE].....travelled to the year 2033, asked the robots (who wiped out mankind).....[/QUOTE] The proper Futuristic expression is "Killed.All.Humans."

[QUOTE=VictordeHolland;353482]travelled to the year 2033 [...] travelled back into time[/QUOTE]

Did you happen to come back for an IBM 5100?

[QUOTE=blahpy;353582]Did you happen to come back for an IBM 5100?[/QUOTE]
They wouldn't let me travel back that far, something to do with the grandfather paradox and alternate timelines.... :-(

[QUOTE=blahpy;353582]Did you happen to come back for an IBM 5100?[/QUOTE]
I just got a permission for time travel, I sent it to myself in two weeks when I will fetch a Datapoint 2200. Is there a prime95 port for it or could I only run factoring work on it? With 8K of serial ram I guess P-1 is out of the question? :smile:

Göran

[QUOTE=Axelsson;353706]I just got a permission for time travel, I sent it to myself in two weeks when I will fetch a Datapoint 2200. Is there a prime95 port for it or could I only run factoring work on it? With 8K of serial ram I guess P-1 is out of the question? :smile:

Göran[/QUOTE]

Huh? I thought you would be [URL="http://en.wikipedia.org/wiki/John_Titor#Outline"]taking it back with you...[/URL]

You're violating the Temporal Prime Directive!

@chalsall:
Any "status" on those LL and DC charts? :smile:

After an extended 'time-off' (holiday, warm weather, etc) I have restarted MISFIT and have rejoined the gpu72 workers.
I shut it down for a long time because I noticed that the fans started to make too much noise.
With a lower temp, they activate less frequently and I can live with the noise it makes now.

Welcome back!

[QUOTE=diamonddave;352216]Is there a reason that the 62M LL range at 73 bit never gets assigned?[/QUOTE]
[QUOTE=garo;352229]I think we should release 62M at 73 bits. People are being assigned LL tests in the 66M range. Clearly we are having trouble keeping the pace. A bit of breathing room is required.[/QUOTE]
[QUOTE=chalsall;352231]Is that an order Commander? My instruments show that things are cool at 74, despite some stray impacts...

Can we wait for Captain Jean-Luc Picard (read: George) for the final authoritative order? The firing solution is programmed... (That's meant to be serious, and funny, at the same time....)[/QUOTE]
[QUOTE=kracker;352233]Then why are we getting 66M exponents for testing? Where is the actual "wave" at now?[/QUOTE]
[QUOTE=Prime95;352245]I do not have a strong opinion. If we are currently assigning 66M LLs that are only TFed to 73, then we should be assigning the 63Ms instead. If the 66M assignments are TFed to 2^74 and Chris believes he has the firepower to both keep ahead of the LL wavefront and get the 63Ms to 2^74, then I'm OK with holding the 63Ms back.[/QUOTE]
[QUOTE=chalsall;352254]Copy. Thank you.
And trust us....[/QUOTE]
[QUOTE=garo;352830]Looks like Primenet is assigning 67Ms. SPE chalsall?[/QUOTE]
[QUOTE=chalsall;352833]Nope.
Instead, very careful management. Note all those assignments are TFed to at least 74, and P-1'ed.[/QUOTE]

Chalsall claims that TF to 74 bits is just keeping pace with LL demand because (for the last few months) every LL assignment has been TFed to 74.
The reason that LLs are assigned in the 67M range is that the TF to 74 effort sprawls that high (for no good reason as far as I can see).
This can't be helping new LL assignments, but there is a fatl flaw in Chalsall's argument which is about to become apparent to all:
Nearly all the exponents being TFed to 74 are already TFed to 73.
The supply of these is approaching exhaustion, and TF from 71 to 74 will not keep pace.
That said, if we settle for TF to 73, we could expect to gradually build up a buffer of about 50,000 exponents between TF and LL assignment.
Then we can change to 74 bits, and hope the 73bit buffer doesn't get used up. I estimate that this moment might occur at exponents around 70M.

By 90M it might be time to go to 75 bits.

There is notime toulouse: cut your losses and TF to 73 now.

David

[QUOTE=davieddy;354002]By 90M it might be time to go to 75 bits.[/QUOTE]

Welcome back David...

We missed you....

:bs meter:

[QUOTE=kracker;354018]bs meter[/QUOTE]

My apologies for this, but I was taught that you should put on your life vest before you put on your boots. And then before you left shore.

Perhaps I was instructed incorrectly...

[QUOTE=chalsall;354029]My apologies for this, but I was taught that you should put on your life vest before you put on your boots. And then before you left shore.

Perhaps I was instructed incorrectly...[/QUOTE]

WHAT? WHO CAAARES?

BOOTS FIRST

Hi, David.

[B][SIZE=2]Plus ça change, plus c'est la même chose.[/SIZE][/B]

Welcome back, David.

Any kids on your lawn, lately?

I think David actually has a valid point. It seems that a breadth-first TF regimen in the upcoming ranges would be preferable to a slowly progressing full depth wave, or at least in addition to the current wavefront style. Might there be interest in doing a breadth-based TF project to take, say, everything from 80M-100M to 69 bits (or 70, even)? We know factors are cheaper at lower bit depths, so why not take advantage of that? I reckon there should be over 44000 factors to take 80M-100M up to the same bit depth as the 50M region, at (just over) half the cost.

Don't like what GPU72 is doing? Anybody is welcomed to take her own exponent from PrimeNet, do TF how deep she likes it, do P-1, then do LL.

[QUOTE=c10ck3r;354115]I think David actually has a valid point.[/QUOTE]

I respectfully disagree. Although I'm happy to debate the issue (again) and be convinced otherwise (so long as the discussion is rational rather than hysterical...).

The stated agenda of GPU to [strike]72[/strike] [strike]73[/strike] 74 is to help find the next Mersenne Prime as quickly as possible.

The best way to do this, in my opinion, is to TF as deeply as Makes Sense according to James' analysis of the cross over point, balanced with our available firepower.

[URL="https://www.gpu72.com/reports/estimated_completion/primenet/"]This report[/URL] shows that we can TF everything up to 67M to 74 "bits" in 229 days, while it's going to take PrimeNet 576 days to LL to that range. Note that we actually don't hold every candidate not yet TFed to 74 -- [URL="https://www.gpu72.com/reports/estimated_completion/"]this report[/URL] shows that it's going to take us 186 days to TF what we actually "own". Note that these reports are updated five times an hour.

Now, if some wanted to do TFing in the 80M to 100M range, I could facilitate that. However, I would argue that work is better left to those CPUs which are still configured to do TFing, and which George is continuing to coordinate.

(Rational) thoughts?

Not all of the GPU72 TF progress is taking place in 62-63M, 73 to 74 bits. I roughly alternate those with lowest available exponent taken from 71 to 74 bits, along with any lower-factored strays that show up.

Somewhere in there I have gotten my Factors-Predicted-vs-Found gap down from around 8, to 2.6 ATM. :smile:

[QUOTE=sonjohan;353983]I shut it down for a long time because I noticed that the fans started to make too much noise.
With a lower temp, they activate less frequently and I can live with the noise it makes now.[/QUOTE]

How old is your GPU?
That increased amount of noise, together with rising temperatures, may indicate some fan overhaul is needed. You may wish to check whether the fans need to be lubricated. Check the fan rpm, if you have the means, that is a good indicator.
I´ve just done that to my GTX560 with success (fan rotation back to normal speed, less noise, and the temperature 15-20 ºC lower).

[QUOTE=lycorn;354182]How old is your GPU?
That increased amount of noise, together with rising temperatures, may indicate some fan overhaul is needed. You may wish to check whether the fans need to be lubricated. Check the fan rpm, if you have the means, that is a good indicator.[/QUOTE]

Indeed.

We sometimes find ourselves in interesting places, don't we?

[QUOTE=chalsall;354123]
The stated agenda of GPU to [strike]72[/strike] [strike]73[/strike] 74 is to help find the next Mersenne Prime as quickly as possible.

The best way to do this, in my opinion, is to TF as deeply as Makes Sense according to James' analysis of the cross over point, balanced with our available firepower.
[/QUOTE]
Well that agenda is certainly what I am talking about. Funnily enough, that's the first time I've heard or read you say that.

James' crossover point is rendered irrelevant because we are firepower limited (especially since Pete went off in a huff of smoke).

However, there is a link between the two, dictated by these simple facts:

1) LL time goes as expo[SUP]2[/SUP]
2) TF goes as expo[SUP]-1[/SUP]
3) Incrementing bit level doubles the time.

It follows that the interval from one bit level to the next occurs when the exponent increases by 2[SUP]1/3[/SUP].

Now if (as I suggested) we could go to 74 bits for exponents>70M, what on earth is the slightest bit ridiculous in the suggestion that 75 bits will be appropriate above 90M?

THIS AIN'T ROCKET SCIENCE.

Try to adjust your intuition to the appropriate time scale, namely 4 years not 4 months.

David

Here is the breakdown.

Do we go bit-first, which means getting one exponent to the optimal bit before even starting the next exponent? Or, do we go breadth-first which means we factor a whole bunch of exponents to one bit, then the next, then the next, etc until we again reach the optimal point? I tend to agree with Chris on this one.

Let's assume we want to factor to 74 bits in the following explanation. Breadth-first SOUNDS like a great idea because if all of a sudden the LL wavefront catches up and the TF work is not sufficiently getting 74-bits work done, we have an entire batch of 73-bits work done, and it really isn't the end of the world if some 73-bits gets done instead of 74-bits. It looks like a very conservative approach.

Bit-first is actually a better idea because if we factor 10,000 exponents from 72 to 73 before then taking all of them to 74, we have spent enough work to instead take 3,333 exponents from 72 to 74. Every single amount of factoring work on an exponent we don't care about NOW is work that isn't useful to us NOW. It's useful to us later, for sure, but we will have more firepower later.

I was interested in working on taking all exponents below 1,000,000,000 to 66 from 65 for a while, until I realized that the LL front will effectively never reach that point and that all my work would be "wasted" as far as the present interests of the project are concerned.

If we put all our efforts into taking 60M-70M to 74 bits instead of doing all this stuff taking 80M to 72 bits or whatever work we're doing on future exponents, we would be leaps and bounds ahead of the LL-wave. It's just that we're doing TF work now that saves LL work in two or three years.

[QUOTE=davieddy;354002]The supply of these is approaching exhaustion, and TF from 71 to 74 will not keep pace.[/QUOTE]
Whether or not it is true, your incessant nagging has chased away a few of the biggest contributors to the TF effort.
I estimate that we would've kept up the TF from 71 to 74 if the big guns were still there.

Could you estimate the pace we would've had if you had <expletive deleted> and left people alone, choosing to do whatever [B]they[/B] want to do?
And yes, we can decide on what we work. Not you, not George, not Chris can decide: I have the possibility to shut down Prime95 or mfaktc whenever I want. Of course, it could be the case that you wish to chase people away, which is the reason you come trolling in this forum.

[QUOTE=davieddy;354199]In a month or so's time it will dawn on you that we need to get expos to 73 bits as a matter of urgency. I have been pointing this out via my broken record since the days before GPUto72 was created.[/QUOTE]

If I remember correctly, you make the same prediction a couple of months ago...

[QUOTE=davieddy;354199]If only Chris hadn't refused to listen to me "because I didn't run a GPU".[/QUOTE]

I refused to listen to you because you didn't, and continue to not, make sense. :sad: