GPU to 66?
 

Hello fellow GIMPers!
I've been looking at the GPU272 group's posts and progress, and started to question the idea. I (think I) understand the desire to push the leading and DC waves to higher bit levels in order to potentially reduce the amount of LL needed. I'm sure the effort is worth it to the LLers and DCers, but is it the best use of the GPUs?
So, I would like to tenatively propose a sister project, GPU to 66, with the intention of pushing exponents between 80,000,000 and 1,000,000,000 to 66 bits, leaving the first 80M to the GPU272 crew. This will consist of taking 2,095,083 exponents from 64 to 65 bits, and 20,415,407 exponents from 65-66 bits, minus any no longer needed due to factors in 64-65.
Anybody with me?
Johannes Schuck

I do not have a GPU to take my comments for what they are worth.

1. The regular TF-LMH will complete the remaining 64-65 bit within a month. And then they will start again at 100,000,000 and go to 66 bits.

2. This will be about a 2-year assignment.....a long time; sort of but not really because it will still be LONG before the LL wave-front will ever get there.

3. I seem to recall reading elsewhere that GPUs preform best at higher bit levels.

mfakt* design means they're more efficient at longer runs, which in general means lower exponents. But higher exponents also need more TF, so they could be long runs too.

As for your idea, the GPU's are pretty much solely responsible for all TF work between 0 and ~60M, and right now we're not caught up. I think this will happen in the next 3-6 months, i.e. within then I think we'll be ahead of the LL wavefront, at which point your idea is worth more consideration IMO. Also go take a look at the LMH forum; there are some people who use GPU's for TF-LMH, although higher than 100M. There is even a special mfaktc version that is more efficient for the short runtimes.

I think it makes more sense to do a depth-first search (i.e. TF as deep as you can for the LL candidates that will be handed out in the next 0-6 month, while largely ignoring the rest), especially given the relatively new, and therefore likely volatile, nature of the GPU/CPU balance and speeds.

In 5 years our GTX5900Ti Extreme "Limited Edition" liquid-nitrogen-cooled GPUs will trial factor all of those in a matter of a few days.

:soapbox:

[QUOTE=Mini-Geek;284007]I think it makes more sense to do a depth-first search (i.e. TF as deep as you can for the LL candidates that will be handed out in the next 0-6 month, while largely ignoring the rest), especially given the relatively new, and therefore likely volatile, nature of the GPU/CPU balance and speeds.[/QUOTE]

Have I told you lately that I love you?
:smile:

David
x

[QUOTE=Mini-Geek;284007]I think it makes more sense to do a depth-first search (i.e. TF as deep as you can for the LL candidates that will be handed out in the next 0-6 month, while largely ignoring the rest), especially given the relatively new, and therefore likely volatile, nature of the GPU/CPU balance and speeds.[/QUOTE]

Isn't that what George started a few months ago when he added 2 or 3 more bit levels in the 50M range?

Or are you proposing more bits yet?

[QUOTE=petrw1;284060]Isn't that what George started a few months ago when he added 2 or 3 more bit levels in the 50M range?

Or are you proposing more bits yet?[/QUOTE]
Yep.
Unfortunately the ~26,000 primenet TF assignments have ground to a halt at 71 bits. The optimal bit level is still governed by the GPU firepower available ATM, but the times they are a changin'.

David

@OP
Bad idea. If you factor all these exponents to higher levels, slower CPUs will not be able to contribute meaningfully to GIMPS. It will be a long long time before even 100M exponents are handed out for LL tests so we need to concentrate on the immediate job first. Make sure every exponent under 60M due to be handed out for LL is factored to 71/72/73 bits.

Sounds like Garo half agrees with me (albeit loathe to admit it).
The brutal fact is that CPUs have been rendered redundant for [B]any[/B] TF.
If the fast guys would let up a bit on DCs (do the man-sized LL work instead), slower CPUs can do them. If the DC wavefront were smaller (say 1/3) than the LL wavefront, the chance of finding a prime per GHzDay would become comparable to that of first time LLs.

David

What do you mean by 'smaller'? I personally put one core to P-1, one to LL, one to DC, and one to mfaktc. That way by numbers I get the most P-1, second most DC, and least LL, while maintaining the same 'GHz-Days' throughput. Unless the DC throughput rate matches the LL throughput rate, I will continue to keep at least one core on DC.

Also, most of us agree in general with garo, and by proxy with you. You just keep spouting off anyways. (We're giving it all she's got, captain!)

[QUOTE=davieddy;284507]The brutal fact is that CPUs have been rendered redundant for [B]any[/B] TF.[/QUOTE]

Everybody agreed with that already. However, we can't go very high with the factoring bit level, as the difficulty doubles for each step. I still do not understand what are you pleading for. Do you want to raise the TF bitlevels higher, or are you arguing that we are already TF-ing too high?

We should raise the bitlevel as long as we still can find factors faster then we would be able to clear the exponents by LL. The percent of factors is "somehow constant" per each bitlevel, but the time to do a bitlevel doubles every step. Right now we should not do any TF higher then 69 at the DC-front (<30M exponent). Even 69 is too much, as we can find a factor every 2 days in average, per GPU involved, and it would take shorter time just to do LL and clear the exponent, using the same GPU. For LL front (55M++ exponents), 72 bit is quite ok we find a factor every 2-3 days, or we could go higher, eventually 73 bits or more, we would still find a factor every 6-7 days in average, per GPU, saving 2 LL tests and some P-1, that would be about 10-12 days or more on the same GPU, saved. But it would make no sense to go higher, if we can't find a factor every 7-8 days maximum, in average, then better do LL directly.

If even me understood that, then it should not be so difficult to understand...

[QUOTE=Xyzzy;284012]In 5 years our GTX5900Ti Extreme "Limited Edition" liquid-nitrogen-cooled GPUs will trial factor all of those in a matter of a few days.[/QUOTE]

A' ye sure they won't be GTX[B]6[/B]90, or [B]7[/B]90? :P

[QUOTE=LaurV;284544]A' ye sure they won't be GTX[B]6[/B]90, or [B]7[/B]90? :P[/QUOTE]

"our GTX5900Ti" .....note extra zero. :smile:
Besides liquid nitrogen cooling, this hypothetical unit may also require its own fusion power plant to keep it cranking.

I noted the extra zero. THAT was the point. Why nVidia should change their habits?

[QUOTE=LaurV;284568]I noted the extra zero. THAT was the point. Why nVidia should change their habits?[/QUOTE]

hyperbole
[URL]http://dictionary.reference.com/browse/hyperbole[/URL]

which is not quite the same as the mathematical planar conic section: hyperbola
[url]http://dictionary.reference.com/browse/hyperbola[/url]

[QUOTE=LaurV;284568]I noted the extra zero. THAT was the point. Why nVidia should change their habits?[/QUOTE]

OK. I just figured that Xyzzy was projecting to a point at which the model numbers had gone up by an order of magnitude. Hence, the facetious reference to fusion power.

[QUOTE=S34960zz;284603]hyperbole
[URL="http://dictionary.reference.com/browse/hyperbole"][/URL][/QUOTE]
[URL="http://dictionary.reference.com/browse/intrudress?qsrc=2446&o=100074"]blearh! [/URL]

[QUOTE=LaurV;284543]Everybody agreed with that already. However, we can't go very high with the factoring bit level, as the difficulty doubles for each step. I still do not understand what are you pleading for. Do you want to raise the TF bitlevels higher, or are you arguing that we are already TF-ing too high?
[/QUOTE]

I mean two things.

1) DCs are about all that's left for the humble CPUs, so eager shithot CPU/GPU owning beavers should just let it ride.

2) If it's worth TFing DCs to N bits, it's worth TFing firstime LLs to N [B]PLUS FOUR.[/B]

Don't be embarrassed if you can't work out 2).
George himself slipped up on that one recently.

POTYmouth

PS and to repeat myself for the umpteenth time,
TFing a 350M expo to 66 bits is like "[B]PISSING INTO THE WIND[/B]"

Talking of getting too high...
...'66 is my favourite musical vintage.
[URL="http://www.youtube.com/watch?v=f4WeqP2G6pI"]Give me an F***[/URL]
What does that spell?

[QUOTE=davieddy;284695]
TFing a 350M expo to 66 bits is like "[B]PISSING INTO THE WIND[/B]"
[/QUOTE]

Right, both tasks are fun.:smile:

[QUOTE=davieddy;284695]
Give me an F**
What does that spell?[/QUOTE]
f-a-c-t-o-r
I thought you know already, you are member of mersenneforum longer then me....
But my factors I keep. They are not for you...
:P

[QUOTE=davieddy;284695]TFing a 350M expo to 66 bits is like "[B]PISSING INTO THE WIND[/B]"
[/QUOTE]

[QUOTE=Chuck;284705]Right, both tasks are fun.:smile:[/QUOTE]

I'm going to PM you to explain how much I treasure this reply.
Brace yourself:smile:

David

[QUOTE=LaurV;284706]f-a-c-t-o-r
I thought you know already, you are member of mersenneforum longer then me....
But my factors I keep. They are not for you...
:P[/QUOTE]

Afraid it was a bit more boring than finding a factor.

It was just [B]FISH.[/B]

David

You haven't lived until you've pissed a wasp all the way to the ground that was trying to "say hello" to your equipment.

[QUOTE=oswald;284715]You haven't lived until you've pissed a wasp all the way to the ground that was trying to "say hello" to your equipment.[/QUOTE]
...

......

:huh:


?

:shock:














































...........That's a pretty big bladder.

:threadhijacked:

Surrealistic Pillow or What?
 

[url=http://www.youtube.com/watch?v=Q1cfTMdjkYM]White Rabbit[/url]
[QUOTE=oswald;284715]You haven't lived until you've pissed a wasp all the way to the ground that was trying to "say hello" to your equipment.[/QUOTE]

[QUOTE=Dubslow;284718]...

......

:huh:


?

:shock:














































...........That's a pretty big bladder.

:threadhijacked:[/QUOTE]

Heh, I just noticed that the 'shock's eyelids twitch.

[QUOTE=Dubslow;284729]Heh, I just noticed that the 'shock's eyelids twitch.[/QUOTE]More to it than that.
If you are referring to Grace Slick, I think she has the finest voice ever, and she was ~30 in 1967. Just ask me, and I'll give you a run down of "must see/hear" US bands of the second half of the 60s. Jefferson Airplane are immortal, but there are several others.

David

Just googled. Born 1939.
Still going strong. Obviously had an easy life and didn't drop enough acid.

IMHO looks to die for.

Nope. Just this: :shock:
Watch the lower eyelids.

[QUOTE=Dubslow;284733]Nope. Just this: :shock:
Watch the lower eyelids.[/QUOTE]
Still not sure what you are on about.
But what I do understand is that she has the capacity to make a man (or woman for that matter) go blind if they look at her too closely.

Seriously, she is the female equivalent of Jim Morrison.

Now you are going to ask me who he was aren't you?

David

I love the links and almost always know what is there before I get there.
I understand the links.
[COLOR=Black][FONT=&quot]Battle not with F** lest ye become a F**; and if you gaze into the link the link gazes into you.[/FONT][/COLOR] (Sorry FN)

But I differ on you should only do a certain kind of work with a certain cpu/gpu. I do the search because I see patterns in the results and it relaxes me. I think people run different kinds of work because they like it.

So, do you enjoy the search or should the search be strict unto itself.

And what happened to CPU to 66?

???
 

[url=http://www.youtube.com/watch?v=4e9CkhBb18E]You talking to me?[/url]

[QUOTE=oswald;284778]
And what happened to CPU to 66?[/QUOTE]
I don't think we've addressed this since post 20.
[QUOTE=oswald;284715]You haven't lived until you've pissed a wasp all the way to the ground that was trying to "say hello" to your equipment.[/QUOTE]I nominate this for POTY 2012. (What an unfortunate acronym in context.)
[QUOTE=davieddy;284738]Still not sure what you are on about.
But what I do understand is that she has the capacity to make a man (or woman for that matter) go blind if they look at her too closely.

Seriously, she is the female equivalent of Jim Morrison.

Now you are going to ask me who he was aren't you?

David[/QUOTE]When I don't understand something in your posts, I typically ignore it. Richard was the exception. And I'm referring to the emoticon :shock: provided by the forum: it moves. One eyelid twitches, then the other, then they both blink, and it does it again.

[QUOTE=Dubslow;284794]I'm referring to the emoticon :shock: provided by the forum: it moves. One eyelid twitches, then the other, then they both blink, and it does it again.[/QUOTE]


I have you know I dreamt about those twitches all night. Now I'm twitching.

:direction:

Services 2 miles?
 

That's good: I'm bursting for a pee.

Perhaps: OP question answered. Thread has now been unharmfully (at least with regards to OP)... re-purposed.

[QUOTE=Dubslow;285000]Perhaps: OP question answered. Thread has now been unharmfully (at least with regards to OP)... re-purposed.[/QUOTE]
"purposed" may be putting it a bit strongly:smile:, but I agree with "nothing further to add as regards the thread title".

BTW rather than overuse the "thread hijacked/diverted" icons, it would be preferable to simply say something new and constructive about the original topic. I think most folk (Bob being a notable exception) are capable of skipping any incidental frivolity.

David