massively scalable, partitioned row store, masterless architecture, linear scale performance, no single points of failure, read/write support across multiple data centers & cloud availability zones. Fast, open source, shared memory (using memory mapped files e.g. Horizontal scalability through sharding, partitioning, replication, and disaster recovery API: Java, REST via (Objective C, Android, etc...), Protocol: Java embedded/Binary Socket/REST, Query Method: Persistence Manager/ORM, Replication: Multicast, Written in: Java, Concurrency: Re-entrant read/write, Misc: Free and Open Source. Query Method: declarative query language (AQL), query by example.API / Query Method: CQL and Thrift, replication: peer-to-peer, written in: Java, Concurrency: tunable consistency, Misc: built-in data compression, Map Reduce support, primary/secondary indexes, security features. Cassandra-compatible column store, with consistent low latency and more transactions per second. in /dev/shm or on SSD), multi process, hash table, e.g. Misc: Linear Scalability, High Availability, WAN Replication, GUI based Administration & Monitoring Tools, Messaging, Runtime Data Sharing, Cache & Item Level Events, Continuous Query & Custom Events, DB Dependencies & Expirations Open Source In-Memory JCache compliant Data Grid. Commercial licensing available Tutorials API: REST, Graph Blueprints, C#, D, Ruby, Python, Java, PHP, Go, Python, etc. Replication: master-slave (m-m to follow), Sharding: automatic and configurable Written in: C/C++/Javascript (V8 integrated), Concurrency: MVCC, tuneable Misc: ACID transactions, microservices framework "Foxx" (Javascript), many indices as secondary, fulltext, geo, hash, Skip-list, capped collections API: REST, Binary Protocol, Java, Node.js, Tinkerpop Blueprints, Python, PHP, Go, Elixir, etc., Schema: Has features of an Object-Database, Document DB, Graph DB and Key-Value DB, Written in: Java, Query Method: SQL, Gremlin, Spar QL, Concurrency: MVCC, tuneable, Indexing: Primary, Secondary, Composite indexes with support for Full-Text and Spatial, Replication: Master-Master + sharding, Misc: Really fast, Lightweight, ACID with recovery.

I've just bough an Asus laptop which is equipped with two graphics cards; one integrated in the intel i3 CPU and a Radeon Mobility HD5145.

Will Ubuntu switch between the cards to balance power/performance? only use the Radeon when the demands placed on the integrated card are too great? If you look in your /var/log/Xorg.0you can see which driver is loaded (page down a few screenfuls and look for RADEON(0) or intel(0).

Right now you can use the GUI provided by Ubuntu Control Center to switch between GPUs, but it only works with the opensource drivers. The only way I know of to force which card to use is to specify the PCI Bus ID of the graphics card. It's also usually shown near the top of your /var/log/Xorg.0

Then, set up your with a device section and put or whatever bus id you want.

» Commercial support Accumulo is based on Big Table and is built on top of Hadoop, Zookeeper, and Thrift. Replication: Master / Slave, Misc: also lists, sets, sorted sets, hashes, queues. It uses memory-mapped files, so it has the read performance of a pure in-memory database while still offering the persistence of standard disk-based databases, and is only limited to the size of the virtual address space, (it is not limited to the size of physical RAM) Sophia is a modern embeddable key-value database designed for a high load environment. Extensive datastore-side-scripting is provided via deeply embedded Lua. API: JDBC, SQL; Wonder DB is fully transactional, distributed New SQL database implemented in java based on relational architectures.

It features improvements on the Big Table design in the form of cell-based access control, improved compression, and a server-side programming mechanism that can modify key/value pairs at various points in the data management process. API: tons of languaes, JSON, Protocol: REST, Query Method: Map Reduce term matching , Scaling: Multiple Masters; Written in: Erlang, Concurrency: eventually consistent (stronger then MVCC via Vector Clocks) API: Tons of languages, Written in: C, Concurrency: in memory and saves asynchronous disk after a defined time. Cheat-Sheet: », great slides » Admin UI » From the Ground up » Fast and web-scale database. Predictable performance; achieves 2.5 M TPS (reads and writes), 99% under 1 ms. Replicated, zero configuration, zero downtime, auto-clustering, rolling upgrades, Cross Datacenter Replication (XDR). APIs: C, C#, Erlang, Go, Java, Libevent, Node, Perl, PHP, Python, Ruby. It has unique architecture that was created as a result of research and rethinking of primary algorithmical constraints, associated with a getting popular Log-file based data structures, such as LSM-tree. So you can get best of both worlds, sql, joins and ease of use from SQL and distribution, replication and sharding from No SQL movement.

(formerly known as Stratosphere) massively parallel & flexible data analytics platform, API: Java, Scala, Query Method: expressive data flows (extended M/R, rich UDFs, iteration support), Data Store: independent (e.g., HDFS, S3, Mongo DB), Written in: Java, License: Apache License V2.0, Misc: good integration with Hadoop stack (HDFS, YARN), source code on Github horizontally and vertically scalable, relational, partitioned row store, document store API / Query Method: SQL (native, DRDA, JDBC, ODBC), Mongo DB wire listener, mixed mode, replication: master / slave, peer-to-peer, sharding, grid operations, written in: C, Concurrency: row, page, table, db locking, Misc: ACID, built-in data compression, scheduler, automatic cyclic storage management, extensible, in memory acceleration, native ports from ARM v6 up Links: Documentation, IIUG, Company. Links: Community, Development, Slides, 2.5M TPS on 1 node at Intel, 1M TPS on Amazon WS, 1M TPS w/ SSDs on 1 Server, Combating Memory Fragmentation API: Many languages, Written in: C, Replication: Master / Slave, Concurrency: MVCC, License: Sleepycat, Berkeley DB Java Edition: API: Java, Written in: Java, Replication: Master / Slave, Concurrency: serializable transaction isolation, License: Sleepycat Oracle No SQL Database is a distributed key-value database. Implemented as a small C-written, BSD-licensed library. Tested performance is over 60K per node with Amazon m3.xlarge VM.

You don't need to specify the video driver as well (but it can't hurt).

David Airlie has worked on making hybrid graphics work better. for 'linux hybrid graphics' may turn up other useful examples and details about it.