error during Jacobi check on 330,000,000+ exponent

[evanh]

should I continue or stop?

[paulunderwood]

Looks okay to me. Prime95 should backtrack to a known good point. Check for dust and heat issues. It could have been a cosmic ray that caused the error!

[evanh]

Originally Posted by paulunderwood

Looks okay to me. Prime95 should backtrack to a known good point. Check for dust and heat issues. It could have been a cosmic ray that caused the error!

Prime95 has been running since I posted this and it is performing a Jacobi Error check after each iteration. I was at 41.76% when I posted this and now at 42.55% running the error check at each iteration print which is every +10,000.

still showing "fair"

is there a log that shows what piece of hardware triggered the error?

I am considering just turning the computer off for tonight. Then starting it back up in the morning allowing everything to sit and cool off.

I run the MONIT widget that tracks system characteristics and it has a thermal indicator similar to an RPM gauge. Blue for cool and red for hot.. its just over halfway probably around 5/8 of the complete red.

[Prime95]

It is NOT running the Jacobi check every print. It is telling you that a Jacobi error check has failed sometime in the past. It will continue to do this for the rest of the primality test.

I personally would not abandon 41% completion because of a single Jacobi failure. There is no way to know what hardware is giving you problems or if a hardware error has gone undetected.

Sometime before you get another work assignment, download the latest prime95 and change you work preference to PRP on 100 million digit numbers. PRP allows for more robust error detection and correction.

[evanh]

Originally Posted by Prime95

It is NOT running the Jacobi check every print. It is telling you that a Jacobi error check has failed sometime in the past. It will continue to do this for the rest of the primality test.

I personally would not abandon 41% completion because of a single Jacobi failure. There is no way to know what hardware is giving you problems or if a hardware error has gone undetected.

Sometime before you get another work assignment, download the latest prime95 and change you work preference to PRP on 100 million digit numbers. PRP allows for more robust error detection and correction.

I definitely want to continue the test, just did not know if the error indicator meant that I needed to take any action.

I did not know that it would print every time so that is good to know that it may not be a continuous item just an occurrence.

I will make the appropriate changes before starting another work assignment.

thank you.

[evanh]

I have successfully updated to version 29.4 build 7 for Mac OSX

I have also changed to 100,000,000 numbers to PRP test.

between my current exponent and my next exponent I will pull a double check. will likely only take a few days.

thank you