Control de usuario
Patrocinadores
Estadísticas
Miembros:
325.587
Online:
1.290
Hilos:
1.351.593
Mensajes:
23.926.300
Stats

Índice de foros PC General

Computación Distribuida en EOL

Para cuestiones informáticas generales: Software propietario, grabación, programas multimedia, etc.

Moderadores: metalgear, coyote

d34th
Avatar de usuario
MegaAdicto!!!
 
Mensajes: 817
Registrado: 08 Ene 2005
Ubicación: Barcelona

Re: Re:

Mensajepor d34th 04 Jul 2009 17:57

jertocvil escribió:
d34th escribió:La verdad es que ahora mismo no se si usa CPU+GPU (que seria lo logico) o sólo gpu
Piensa que en caso de que solo use GPU, siempre puedes poner a ejecutar un cliente CPU al mismo tiempo

Por cierto, con que grafica tienes pensado procesar?

Con una GeForce 9400GT. Si que es compatible con CUDA. Aunque no se que pasa que cuando pongo el cliente de GPU empieza ser inusable todo el PC.


El problema es que con GPUs de gama baja, pasan estas cosas, yo tengo una 8600 y me va todo bastante mal cuendo lo pongo
Para ir bien, lo deseable seria, minimo, una 8800, o una 9800
ImagenImagen

Rolod
Avatar de usuario
Fission Mailed
 
Mensajes: 2564
Registrado: 19 May 2005
Ubicación: Unión Europea

Mensajepor Rolod 27 Ago 2009 18:12

Completada la 1ª fase de "Descubriendo fármacos contra el Dengue"

“Discovering Dengue Drugs - Phase 1” – DONE!

All Phase 1 results that you have generated are now on our disk servers where they are being compiled, processed and double-checked. Thank you for your efforts and help!

Phase 1 is only the first step on our path to discovering dengue (and West Nile, hepatitis C, yellow fever, and Leishmania) drugs. Phase 1 has systematically tested ~3 million small "drug-like" molecules against 18 different protein targets (12 viral protease targets, 6 Leishmania enzymes). These 54 million individual docking experiments were completed in less than 2 years, even with the devastating interruptions and slowdowns wrought by Hurricane Ike. Our 30-node high-performance cluster would have needed to run nonstop for 205 years to finish these calculations.

Phase 1 identified several thousand molecules that could interact with each target protein. Typically, ~90-95% of Phase 1 “hits” turn out to be false positives when tested in the laboratory. Thus, Phase 1 hits are very inefficient to pursue with laboratory testing.

“Discovering Dengue Drugs - Phase 2” is designed to reduce the number of Phase 1 false positives. In Phase 2, we will complete ~2000 computationally demanding free energy calculations for each protein target (each calculation uses a promising Phase 1 hit). These calculations will eliminate many of the Phase 1 false positives and produce refined hit lists that are expected to contain ~80% true positives. Testing Phase 2 hits in the laboratory will be much more productive, efficient, and rewarding than testing Phase 1 hits. For example, to find 25 small molecules that stop dengue virus replication in the laboratory, we would need to synthesis and test either 250-500 Phase 1 hits or ~30 Phase 2 hits. We like the improved odds provided by the Phase 2 calculations.

Alpha testing has begun for “Discovering Dengue Drugs – Phase 2”. This work is moving along remarkably smoothly (special thanks to the IBM team of professionals). Beta testing and Phase 2 will launch very very shortly.

I look forward to your continued help and generosity as we work together to find cures for dengue, West Nile, hepatitis C, yellow fever, and Leishmania diseases.

All my best, and many thanks!

Stan coffee

P.S. Although completed, DDDT-1 will remain officially active on the grid. From time to time, we will need to test additional drug targets to help refine our evolving drug predictions, so do not be surprised if new DDDT-1 workunits show up in your cache.

¡Felicidades a todos los que hayáis participado!

Fuente
Imagen

Imagen Imagen

Retroceder, nunca; rendirse, jamás.

Previo

Volver a General

¿Quién está conectado?

Usuarios navegando por este foro: No hay usuarios registrados visitando el foro y 0 invitados