Hi everyone,
We've had a new CUDA app in beta for around the last 24 hrs and it seems stable enough to push it out of beta at this time.
Here's the technical explanation from the lead developer of the app, MC "PseudoGravity":
While each Minecraft world possesses a single world seed, every 16x16 chunk within the world possesses its own derived seed, known as a decoration seed. Each chunk’s decoration seed is determined by the unique combination of world seed and chunk location. This decoration seed in turn guides the generation of features such as trees, ores, and portions of structures.
The decoration seed formula is as follows:
decorationSeed = (AX + BZ) ^ worldSeed
Here, A and B are 64-bit pseudorandom odd numbers generated from the world seed and are therefore the same throughout the world. Variables X and Z are the coordinates of the chunk and are always multiples of 16. This formula is designed to ensure that each chunk within the world has a different decoration seed and thus different trees and related features.
However….
Taking a closer look at this equation reveals that the portion AX+BZ is the only part that depends on the chunk coordinates. If this subexpression produces the same value at multiple locations within the same world, then those chunks will have the same decoration seed. Specifically, if NA=MB mod 2**64 for some small integers N and M, then
A(X+N) + B(Z-M) = AX+BZ mod 2**64
and we see that the decorations seeds repeat along a line formed by (N,-M). When N and M are very small, interesting effects occur, as chunks form stripes along which identical trees and related features repeat endlessly.
Participants in this GPU-based BOINC project will exhaustively comb through all 2**64 Minecraft world seeds to uncover those in which the pseudorandom values A and B produce repeating patterns. Preliminary searches have discovered a few diagonally-repeating worlds, including several which repeat every single chunk diagonally. This full-scale search will produce a complete list of all Minecraft worlds in which decoration seeds repeat within a distance of 8 chunks, including non-45-degree angles.
View article · Sun, 27 Jul 2025 03:42:13 +0000
Aliquot sequence 3326040 has terminated!!!
View article · Sat, 26 Jul 2025 06:24:14 +0000
Dear volunteers,
Apologies for the delay in getting back to you since the simulations were completed. Over the past few weeks, and with a special mention to all of you, Iván defended his Final Degree Project. In the weeks leading up to it, we were somewhat disconnected as we reviewed his project and prepared for the defense. He achieved an excellent grade (one of the highest in his class). Congratulations to Iván for all the hard work!
These days, we are preparing a paper for the Spanish Society of Biomedical Engineering Conference, to present the new features of DENIS and some initial performance tests. The idea is to later carry out a larger set of tests to see how volunteer computing can help with this type of problems, which types of simulations could benefit most from your support, and what the limitations are.
Once this more technical part is further along, we have two challenges ahead. The first, which is delayed on my part, is to run a few more simulations for the model optimization and publish all the results. The second is to advance in the area of model optimization, but now using the fiber simulator: this opens many new doors for us.
The holiday season is approaching, so I’m not sure how much we’ll be able to launch before then, but we’ll keep you informed as things progress.
Thank you all very much for your support.
Best regards,
Jesús
============================================================
Estimados voluntarios:
Disculpad por haber tardado un poco en daros noticias desde que se terminaron las simulaciones. En las últimas semanas y con una mención especial a todos vosotros, Iván defendió su Proyecto Final de Grado. En las semanas previas estuvimos algo desconectados revisando su memoria y preparando la defensa. Consiguió una magnífica calificación (de las mas altas de su promoción). ¡Enhorabuena Iván por todo el trabajo realizado!
En estos días estamos preparando un artículo para la Conferencia de la Sociedad Española de Ingeniería Biomédica, para presentar las nuevas funcionalidades de DENIS y unas primeras pruebas de rendimiento. La idea es hacer más adelante una batería de pruebas más grandes para ver cómo la computación voluntaria puede ayudar en este tipo de problemas, qué tipos de simulaciones son las que más se podrían beneficiar de vuestro apoyo, y cuáles las limitaciones.
Una vez avanzada esta parte más técnica, tenemos dos retos por delante. El primero, que van con retraso por mi parte, es hacer algunas simulaciones más de la optimización de modelos, y publicar los resultados. El segundo es avanzar en el tema de la optimización de modelos, pero haciendo uso ya del simulador de fibra: aquí se nos abren muchas puertas nuevas.
Se acercan ya los días de descanso, por lo que no sé cuánto podremos lanzar antes, os iremos informando según vayan saliendo.
Muchas gracias a todos por vuestro apoyo.
Un saludo,
Jesús.
View article · Mon, 21 Jul 2025 08:43:31 +0000
Computing power · Certificate · Poll
Programmers:
BOINC client 8.2.4 released
The 8.2.4 version of the BOINC client has been released for all platforms. Download it here. Release notes are here. The new version supports Docker apps - an important step for BOINC.
10 Jul 2025, 21:46:02 UTC
· Discuss
Lines of code visualization
Vitalii made a visualization of the number of lines of code in BOINC going back to 2002.
16 Mar 2025, 9:07:20 UTC
· Discuss
Grafana project dashboards
Check out Grafana project dashboards showing time-varying graphs of project info such as number of unsent and in-progress jobs.
15 Feb 2025, 21:03:13 UTC
· Discuss
Copyright © 2025 University of California.
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License,
Version 1.2 or any later version published by the Free Software Foundation.