Hello, from another newbie

[kriesel]

Originally Posted by kladner

v30.3b6:

I don't see PRP as an option, except for PRP for Mersenne cofactors. Am I correct in thinking that selecting Whatever, 1st Time, or DC will yield some kind of PRP assignment?

First time will give lowest available first time test exponents.
World record will give exponents available above the largest known Mersenne prime.
Sometimes those are different, but not now. Either would get assignments around 100M exponent.
100,000,000 digit will give exponents above 332 million. (which will take about 11 times as long.)

In manual checkout you need to choose between PRP and LL.
prime95 v30.3b6 will usually get PRP work. Earlier versions gave separate choices in prime95 for PRP or LL in each exponent category. Since implementation of PRP proof at half the total effort of first test and double check, there's a strong incentive to move away from LL to PRP for initial testing.
Some history here for the concept and here for prime95 v30.3

If and when a probable prime is found, it will need to be confirmed prime by the definitive LL test performed with more than one program on more than one set of hardware by multiple people. During the confidentiality period.

Found a new prime? http://www.mersenneforum.org/showpos...5&postcount=14
Verification of a new find https://www.mersenneforum.org/showpo...83&postcount=8

[mathematizer]

Originally Posted by moebius

for v.6.11.64 -6.11.80 you only need uploader.exe attached in proofuploader.zip

windows 10

cd proof (my preferred directory for uploader.exe)
for %a in ( *.proof ) do uploader your-primenet-uid %a

I can't give a guarantee for v7, better if you contact user preda the developer of gpuowl, because v7 uses proofs v2

Your reply zipped right over my head. Sorry!

I’m accustomed to copying my results from the file into the manual upload section. I realize now that that practice was not preferred..

I believe you are describing a method to upload proof of my (likely Composite) results using a previous version of gpuowl and suggesting I contact the creator for clarity on how to upload my proof v2 results, if I stick with the current version that I’m using? Presumably to save the need for a doublecheck?

[kladner]

Originally Posted by moebius

At least you will get $ 3000 Awared if you can track down a prime number with PRP.
But rather Mr. Woltman should comment on this.

Thanks for the answers. My interest in asking was aimed at getting PRP, with a mild preference for DC. So, off to study prime, local, and undoc some more.

[moebius]

Originally Posted by mathematizer

Your reply zipped right over my head. Sorry!

I’m accustomed to copying my results from the file into the manual upload section. I realize now that that practice was not preferred..

I believe you are describing a method to upload proof of my (likely Composite) results using a previous version of gpuowl and suggesting I contact the creator for clarity on how to upload my proof v2 results, if I stick with the current version that I’m using? Presumably to save the need for a doublecheck?

I only use gpu owl. v6.11.64 -6.11.80, simply because it has shorter iteration times. I added a method to my original post to upload any .proof file.Thank you, I think you understood me very well.

[Viliam Furik]

Originally Posted by mathematizer

...with Prime95 running, it's 29C.

Are you SURE that is right? If Prime95 is running on all cores, it should be pretty hot (> 50 °C). My Ryzen 9 3900X (12 cores) is running at about 65-70 °C (with AIO) when running Prime95 on all 12 cores, non-multi-threaded. Also, the 9th generation of Intel is known to be hot.

[mathematizer]

That’s the value as measured by the AIO software. I’ll double check tonight.

[kruoli]

Maybe that's the temperature of the coolant.

[Viliam Furik]

Originally Posted by kruoli

Maybe that's the temperature of the coolant.

That would also seem too low. My coolant is about 40 °C, at full load (NZXT Kraken X72).

---
EDIT: Only way I see it to be possible is that it is the idle temperature, assuming it's a correct temperature. If assuming incorrectness, it can be user error, HW or SW bug/error, or some unknown problem.

[mathematizer]

I remember seeing corsairs AIO iHue software reporting 29.1 and 29.4 last night on the CPU (Intel i5) running v30 b6 last night. But I’ll look closer at it tonight.

[Viliam Furik]

Originally Posted by mathematizer

I remember seeing corsairs AIO iHue software reporting 29.1 and 29.4 last night on the CPU (Intel i5) running v30 b6 last night. But I’ll look closer at it tonight.

Please post screenshots of the software.

[mathematizer]

Originally Posted by Viliam Furik

Please post screenshots of the software.

Just got home from work. Here's the screenshot where I got the temperature from. Might be that of the coolant? It’s a few degrees higher than what I posted last night, but it’s been running for longer. Prime95 is currently running and has been doing so for the last day. Looks like it has generated 3 Gerbics/double-check errors during that time, but still has excellent confidence. Wish there were zero errors. It’s about 10% through the PRP.


Also, saw a spot in the AIO software (took a screenshot of that too) where I can enable sensor logging - would that be a good place to gather useful data?