Actualités en direct des projets (non traduites)

Détails

Hi everyone,

We've added some new binaries for this project to allow macs to run.
Both x86 and ARM64 based macs should run just fine.

The linux and windows binaries remain at v1.05, there'd be nothing new for them if we did bump them to 1.06.
Macos is on 1.06 because I uncovered a bug in my deployment of the macos binaries and had to use a version bump to replace the old version.

Anyway, something interesting I've discovered is that M1 (and really, all apple silicon so far) LOVES these tasks.

On a sample mini task, I got 120 seconds on my 12700k, and 61 seconds on my M1 macbook air.
On the same sample mini task, one of our project developers tried it on their M4 Pro and got 43 seconds.
Of course, that's not perfectly reflective of runtimes, but I expect to see apple silicon based machines top the charts for this app based on these preliminary results.

Happy crunching, everyone!

Source

Création : 23 avril 2025

Détails

Hi everyone,

I've added the following new badges while I was doing some database work today.

1. 1.21 Panorama badge. It's a copper bulb block from Minecraft, which is a block found in Trial Chambers. The pano121 app attempted to find a seed which produces a particular trial chamber matching the one found in the title screen panorama for 1.21.
2. Xoroshigo v1.0. This art was created largely by the lead developer of the xoroshigo project, but I added the "alpha" symbol at the bottom right to represent that it was the first run of the app before optimizations were made.
3. Xoroshigo v1.1. Same art as v1.0, but without the alpha symbol.
4. Xoroshigo Infinity. Same art as v1.0 and v1.1, but with an infinity symbol. I attempted to identify everyone who submitted a task after deadline during the time we had a broken version of the app running that caused runtimes measuring in multiple days for some configurations. If you believe I missed you in error, please let me know. We're in the middle of optimizing the amount of data we're storing in our database, so I may need to review previous versions of the data to ensure I don't miss anybody.

I also managed to re-run the badge assignment for the tall cactus project from years ago. Around 2300 or so participants gained the badge, congratulations!
The discrepancy was likely caused by the method that the previous administrator before me used to apply the badge. Since it did not use a sql clause, it likely used a "snapshot" of data without keeping it updated.

All that said, I will probably also add a loneliest seed badge tonight before I wrap up, but I'll update here if I do.

Thanks for your patience as we've been working diligently to get things running smoothly here.

Source

Création : 21 avril 2025

Détails

On 8 April 2025, 12:49:49 UTC, PrimeGrid’s Sierpiński/Riesel Base 5 Problem project eliminated k=67612 by finding the mega prime: 67612*5 ^5501582 +1 The prime is 3,845,446 digits long and enters “The Largest Known Primes Database” ranked 92^nd overall. 27 k’s now remain in the Sierpiński Base 5 problem. The discovery was made by Kai Presler (Aperture_Science_Innovators) of Australia using 8 cores of an AMD Ryzen 9 7945HX with 14GB RAM, running Linux Mint 21.3. This computer took about 1 hour, 24 minutes to complete the probable prime (PRP) test using PRST. Kai is a member of team [H]ard|OCP. The PRP was confirmed prime on 8 April 2025, 20:19:10 UTC, by an AMD Ryzen 9 7950X3D @ 4.20GHz with 128GB RAM, running Debian 12.5. This computer, using 4 cores, took 3 hours and 44 minutes to complete the primality test using LLR2. For more details, please see the official announcement.

Source

Création : 21 avril 2025

Détails

On 3 March 2025, 07:53:17 UTC, PrimeGrid's Generalized Fermat Prime Search found the Mega Prime: 13427472 ⁵²⁴²⁸⁸ +1 The prime is 3,737,122 digits long and will enter “The Largest Known Primes Database” ranked 15th for Generalized Fermat primes and 94th overall. The discovery was made by Jean-Luc Garambois ([AF>Amis des Lapins] Jean-Luc) of France using an NVIDIA GeForce RTX 4080 SUPER in an AMD Ryzen Threadripper 3990X 64-Core Processor with 256GB RAM, running Linux Ubuntu 22.04.5 LTS. This computer took about 15 minutes and 23 seconds to complete the probable prime (PRP) test using Genefer22. Jean-Luc is a member of the L'Alliance Francophone team. The PRP was confirmed prime on 17 April 2025 by an AMD Ryzen 9 7950X3D @ 4.20GHz with 128GB RAM, running Debian 12.5. This computer took about 20 hours, 35 minutes to complete the primality test using LLR. For more details, please see the official announcement.

Source

Création : 21 avril 2025

Détails

Hi everyone,

I appreciate your patience while we worked through this.

v1.05 addresses the extreme runtime issues we've been facing lately.

I would recommend cancelling any v1.04 tasks that are running too long.

I'm looking into ways we can credit those who attempted to run broken v1.04 tasks, but as far as I can tell, boinc does not support that scenario. So I don't have an answer quite yet.

I'm deploying some more tasks worth 10,000 credits each as a consolation for this.

Please let me know if you run into any issues.

Source

Création : 18 avril 2025

Détails

Upgrade shouldn't take too long, but just wanted everyone to be aware.

Source

Création : 18 avril 2025