I don't understand how someone could reserve more than two thousand exponents in the category 1 double checking range. Especially since the reservation rules do not seem to be applied to trial factoring. It was not done through the normal manual reservation page.

Good thing is that the assignments are slowly being completed.

Jacob

[QUOTE=S485122;483618]I don't understand how someone could reserve more than two thousand exponents in the category 1 double checking range.

Jacob[/QUOTE]

You know about this address?

[url]https://www.mersenne.org/manual_gpu_assignment/[/url]

[QUOTE=masser;483653]You know about this address?

[url]https://www.mersenne.org/manual_gpu_assignment/[/url][/QUOTE]I remembered that address, but the assignments do not seem manual assignments. Anyway at that page one cannot reserve exponents in the double-check range at the moment. ("Error code: 40 / Error text: No assignment available for GPU trial factoring, cpu_id: 1455, user_id = 2"

I think the reservation was made by having the workunits in a wortodo file and communicating with the server and that should be refused in at least the category 1 range.

Jacob

[QUOTE=S485122;483661]I remembered that address, but the assignments do not seem manual assignments. Anyway at that page one cannot reserve exponents in the double-check range at the moment. ("Error code: 40 / Error text: No assignment available for GPU trial factoring, cpu_id: 1455, user_id = 2"

I think the reservation was made by having the workunits in a wortodo file and communicating with the server and that should be refused in at least the category 1 range.

Jacob[/QUOTE]

Have you been to [url]www.gpu72.com?[/url]

Its a side-project of GIMPS.
They reserve large chunks of DCTF or LLTF work and dole them out to anyone on the site.

[QUOTE=petrw1;483667]Have you been to [url]www.gpu72.com?[/url]

Its a side-project of GIMPS.
They reserve large chunks of DCTF or LLTF work and dole them out to anyone on the site.[/QUOTE]They may have made the assignments. But I think it should not be possible to reserve in the Category 1 range unless by conforming to the assignment rules. And clearly those assignments do not respect the rules. There is enough TF work outside of Category 1.

Jacob

[QUOTE=masser;483653]You know about this address?

[url]https://www.mersenne.org/manual_gpu_assignment/[/url][/QUOTE]

I didn't, but thanks, it just made it a lot easier for me to reserve new work in the future for my GPU :smile:

[QUOTE=KEP;483675]I didn't, but thanks, it just made it a lot easier for me to reserve new work in the future for my GPU :smile:[/QUOTE]

Or if you want to focus your GPU-TF efforts on the LL or DC leading edges consider this:

[url]www.gpu72.com[/url]

It's a thing Chris is working on. He's using my GPUs to squeeze out another bit level before they get assigned for DC. My turn around time on DCTF is about 2 days and I'm DCTFing about 70 per day.

[QUOTE=rudy235;483519]At last! A reasoned explanation.[/QUOTE]

Also, bear in mind that in some cases, multiple expiration rules may actually be in play at the same time, so when I came up with the query to give the time-to-expire, I ended up using the value that expires sooner if it's under 10 days (I think).

For instance, you may have 30 days to complete but if you haven't started it yet, the expiration time is shorter. But yeah, once you check in even 0.1% of progress it counts as "started" and then the other rules are in play.

It's been a while since I looked at it... I think there was some other case where it could expire sooner, but I don't remember what (or I'm thinking of something else) :smile:

[QUOTE=petrw1;483679]Or if you want to focus your GPU-TF efforts on the LL or DC leading edges consider this:

[url]www.gpu72.com[/url][/QUOTE]

Well I'll see in about 170 days when my ~11700 tasks running from 68 bit to 72 bit completes. I'm not having the fastest GPU in the ballpark anymore, so I'm not really sure that the 86.4 GHz days produced each day (n=689M) is going to make much of a difference on the push to 1 bit higher effort. It is however going to bring us ~720 candidates closer towards the 20,000,000 unfactored candidates milestone :smile:

[QUOTE=KEP;483754] It is however going to bring us ~720 candidates closer towards the 20,000,000 unfactored candidates milestone :smile:[/QUOTE]

:party: