Grid computing and petecomputing have common roots in efforts
aiming at overcoming limitations of supercomputing. This talk,
based on the research in the framework of the European Grid
projects presents some results which potentiall may be
adapted in petacomputing for running large multiscale applications
in interactive way, as well as for development of new programming
paradigms and tools.

First, we present a unified architecture of Grid services
and a tool environment for interactive compute- and data-intensive
applications together with a tool for online, interactive
performance analysis (G-PM) combined with online monitoring
of Grid applications (OCM-G).

Often large scale applications, which combine both time and event driven
simulation may be coupled together with HLA and we have elaborated
an integrated framework (G-HLAM) for running High Level
Architecture (HLA) distributed large-scale simulations on the Grid.

Our solution of abstract workflow composition in an semantic Grid
environment as well as a concept and prototype of a distributed, scalable,
semantic-based and Grid-enabled registry for storing information about
Grid services may be considered as an approach to programming
complex application for petascale systems. This workflow system
is accompanied by a P2P tool for monitoring of resources and
applications (GEMINI).

Distributed component approach is an interesting
alternative to service oriented architectures and we have
elaborated the MOCCA distributed component framework.

Our  methodology for porting legacy code to
Grid systems based on service-oriented architectures
(the LGF system) may be also re-used in petacomputing.

Finally, we present the Virtual Laboratory which will
provide the medical doctors a decision support system to rank
drugs targeted at patients, and the virologists -  an
advanced environment to study trends on an individual, population
and epidemiological level.