Please Note: This workshop is not open to the general public, but only to active researchers.

Registration is now closed.  We will be live streaming the talks. 

Questions can be sent to the workshop email address: [email protected] .

Organizers: Anatoly Spitkovsky (Princeton University), Frederico Fiuza (SLAC/Stanford University), Benjamin Crinquand (Princeton University), Jens Mahlmann (Princeton University), Arno Vanthieghem (Princeton University/NINS) 

Numerical simulations play a leading role in understanding the most efficient particle accelerators in extreme astrophysical environments, such as neutron stars and black holes. Despite the fast progress in the last decade, they prove particularly challenging. Plasmas in the vicinity of compact objects are quintessential examples of complex physical systems that exhibit the interplay of nonlinear processes and many-body phenomena across very disparate scales. Plasma micro-physics induce non-thermal particle distributions in accreting compact objects or blast waves, but multimessenger observational signatures are also affected by the global dynamics. Accurately resolving the separation of scales expected in nature from first principles will not be possible even on exascale supercomputers. The development of novel methods that accelerate simulations is thus particularly timely. Guided by theoretical arguments, machine learning techniques and hybrid methods can exploit increasingly abundant data and high-fidelity large-scale simulations to develop better sub-grid descriptions. This workshop aims at bringing together experts in theoretical plasma physics and astrophysics, numerical methods, and machine learning to discuss new pathways to an improved multiscale modeling of plasmas and a better connection to observations.”

REGISTRATION REQUIRED but is now closed.