Features

General	Multiple Transports
Transactions Cache service Logging service Management	Support for Java Message Service Cajo
Flexible Architecture	Security
Default Environment Initialization Dynamic Code Execution Prioritization	DAC sessions

Comparison with other frameworks

Feature/Framework	GridGain	Hadoop	Hazelcast	Jade	DAC
Map/Reduce paradigm	Yes	Yes	Yes	No	Yes
Load balancing and scheduling	Yes	Yes	Yes	Yes	Yes
JMX-based Management & Monitoring	Yes	Yes	Yes	(1)	Yes
Fail-safe	Yes	Yes	Yes	No	Yes
Zero deployment model	Yes	Yes	Yes	Yes	Yes
Hybrid Cloud	Yes	Yes	Yes	Yes	Yes
Multiple nodes in one VM	Yes	Yes	Yes	No	Yes
JEE integration	Yes	No	Yes	No	No

(1) uses jademx - an open source implementation of Java JMX

Map/Reduce paradigm: Allows developers to write programs that process massive amounts of unstructured data in parallel across a distributed cluster of processors or stand-alone computers.
Load balancing and scheduling: Technique to distribute workload evenly across agents, in order to minimize response time, and avoid overload.
JMX-based Management & Monitoring: All major monitoring and statistical information about all nodes in the grid available via programmatic interface as well as through any JMX-compliant Web or standalone GUI viewer such as JConsole.
Fail-safe: During computation the tasks of failed node(s) are re-distributed among the remaining nodes.
Zero deployment model: Allows to start multiple nodes on the same computer without any changes in configuration.
Hybrid Cloud: Environment in which some resources are provided and managed in-house and others provided externally.
Multiple nodes in one VM: Allows to start multiple nodes in the same Vrtual Machine.
JEE integration: Allows frameworks to be integrated into the application servers.

Cloud Aware Communication & Deployment - allows GridGain runtime to adapt to different environment without affecting any business logic that is running on it
Annotation-based grid-enabling with AOP - allows to grid-enable existing code without any modification by using Java annotations and AOP-style (Aspect-oriented programming) crosscutting effectively forming a grid-enabling DSL (Domain Specific Language)

File system - provides a distributed file system (HDFS) that stores data on the compute nodes, providing very high aggregate bandwidth across the cluster
Master/slave architecture of Map/Reduce framework - a single master server or jobtracker and several slave servers or tasktrackers, one per node in the cluster
Master/slave architecture of HDFS - a single Namenode, master server that manages the filesystem namespace and regulates access to files by clients and a number of Datanodes, one per node in the cluster

Distributed implementations of java.util.{Queue, Set, List, Map}
Distributed implementation of java.util.concurrency.locks.Lock
Distributed implementation of java.util.concurrent.ExecutorService
Distributed MultiMap - for one-to-many relationships
Distributed Topic - for publish/subscribe messaging
Socket level encryption - for secure clusters
Persistence - synchronous (write-through) and asynchronous (write-behind) persistence
Second level cache provider - for Hibernate
Sessions - dynamic HTTP session clustering
Information - support for cluster info and membership events
Dynamic discovery - solved by multicasting
Data management - dynamic partitioning with backups

Compliant with the FIPA specifications - for interoperable intelligent multi-agent systems
Agent communication - Peer-to-Peer intelligent agent communication performed through FIPA ACL (Agent Communication Language) messages
API - extensive set of API’s, alongside with an excellent suite of programming and testing tools
GUI - graphical user interface to manage several agents and agent containers from a remote host
Transport mechanism - adapts to each situation, by transparently choosing the best available protocol (currently: Java RMI, event-notification, HTTP, and IIOP)
Protocols - more protocols can be added via the MTP and IMTP JADE interfaces
JessBehaviour - allows full integration with JESS rule engine
Secure work - allows to work in a proactive way, according to usage rules given by owners
Coordination - allows to coordinate in order to solve complex problems in a distributed way

Hello World - demonstrates direct grid task execution
Distributed JUnit - demonstrate how GridGain can distribute JUnit tests or test suites across grid
MapReduce with Prime Numbers - demonstrates how prime number check it can be grid-enabled using GridGain
Resize Images - demonstrates how to scale and convert images on the grid using JDK encoders
JBoss Cache Affinity Split Example - shows how with GridGain and JBoss Cache you can co-locate data access with your processing logic
MapReduce with Node Metrics - jobs will be routed to the nodes that have at least 2 physical CPUs and average idle time more than 50%
Grid Task Events Processing - being deployed/executed task produces number of events which can be queried later to analyze
Grid Discovery Events Processing - during startup new node communicates with remote ones to set up a connection and exchange attributes
Grid Direct Task Deployment - explicitly deploy a task
Resource injection - work with some storage (or with database the same way)
Monte-Carlo - calculates a risk value for a portfolio of loans

WordCount - reads text files and counts how often words occur
Python WordCount - WordCount example translated into Python and translated using Jython into a Java jar file
C/C++ WordCount - WordCount example using C/C++
Grep - extracts matching strings from text files and counts how many time they occured
Sort - simply uses the map/reduce framework to sort the input directory into the output directory

Hello World - shows a minimal agent printing "Hello World!" and then terminating
Behaviours - includes examples showing the usage of simple and composite behaviours
Messaging - includes examples showing how to make JADE agents communicate
Yellow Pages - includes examples showing how to use the Yellow Pages service provided by the DF agent
Topic-based communication - includes examples showing how to use the topic-based message delivery mechanism introduced with release 3.5
Book Trading - shows a very simple multi agent system for trading books online
Protocols - shows the usage of the support provided by JADE for interaction protocols
Ontology - shows how to use the support for application specific ontologies
In Process - shows the usage of the jade.wrapper package to start a JADE runtime and control a number of agents from an external application
Base 64 - shows how to encode serialized Java objects into the content of an ACLMessage; it also shows how to force the platform to use bit-efficient and/or XML ACLCodec
Mobile - an example of a mobile agent