Database Landscape Map – December 2012

As previously mentioned, one of my most popular pieces of research while at 451 has been the database landscape graphic we produced for our NoSQL, NewSQL and Beyond report.

I recently published an updated version but noted that there were a group of database vendors that had emerged in 2012 that didn’t easily fit into the segments we’d created.

In order to address that I went back to the drawing board and, taking inspiration from London Underground and The Real Story Group, set about mapping the connections between the various players in the database space.

Note: the latest update to the map is available here.

I’ll be honest – I’m not convinced that this is as practically useful as the original, although I believe it is more accurate and it was an exhausting interesting exercise to put it together.

If anyone spots any glaring omissions or errors please keep them to yourself let us know. Additionally, the image is also available on posters, mugs, t-shirts and mouse pads, for a small fee 🙂

Of course, if you’re looking for some perspective on what this all means, I can recommend one of our highly competitive subscription packages

Tags: , , , ,


#1 Stephan Tual on 12.20.12 at 7:08 am

This is absolutely brilliant Matthew. The ‘extensions’ lines are particularly clever. Thank you for putting so much effort in.

#2 Ken on 12.20.12 at 5:48 pm

Nice work Matt! This is a massive undertaking…

#3 Matt on 12.21.12 at 12:35 am

Great stuff, thanks for the effort! I noticed you are missing some in-memory databases though (Oracle TimesTen, IBM solidDB to name a couple)

#4 Matthew Aslett on 12.21.12 at 5:20 am

Thanks Matt, I’ll try and figure out where to put them on for the next rev, and whether “in-memory” deserves its own line

#5 Matthew Aslett on 12.21.12 at 5:21 am

To add:
– SQLite
– TimesTen
– solidDB
– MonetDB
– move SciDB
– Adabas
– in-memory line?
– spatial line?
– TempoDB
– reflect BerkeleyDB’s SQL API
– UniData/UniVerse
– Oracle DBA (integrated with Cloudera)
– WakandaDB
– VertiCloud
– Amazon Elasticache

#6 Matt (a third) on 12.21.12 at 10:00 am

Very cool. RRD and ESE?

#7 @covoutech on 12.21.12 at 11:52 am

To Add:

#8 Matthew Aslett on 12.21.12 at 12:05 pm

GemStone lives on in GemFire and SQLFire, both on the map

#9 mariuz on 12.23.12 at 1:55 am

You forgot Firebird , One of the best MySQL/Oracle alternatives

#10 Alhabshi3k on 12.28.12 at 3:13 pm

a couple of comments on the landscape to add:
– addition to mention DBs : MonetDB still compete with PostgreSQL and SciDB.
– For the database classes/types, to consider the following:
— Semantic Database (AKA Graph RDF) that are actually branched from Graph DB and provide some of the semantic web definitions and SPARQL. such OpenLink Virtuoso, Strabon .
— Geospatial database, that support geographic datatype. a lot of the mentioned DBs in landscape support them (postgresql/postGIS,SQL server,MangoDB, couchDB , Neo4j, spatialite, Strabon,MonetDB)
— Serverless database, that are not running as server and can be file-base or in-memory. such SQLite and spatialite.
– Regarding the licensing types, it would be better to know the licenses using highlighting (color, bold, italic, underline) ; and the license between braces f=free,os= open source and c=commercial. Like Neo4j (OS/C) SQL server (f/c), MonetDB (OS)

#11 Matthew Aslett on 01.02.13 at 5:03 am

Thanks Alhabshi3k,

I was thinking for branching the semantic DBS but it’s already a bit crowded in there. I’ll take a look for another rev. The licensing info is a great idea but again is likely to be hard to squeeze on there. I also looked at geospatial but as you suggest it crosses a number of different database types. Perhaps could be another ‘line’.

#12 Mar4uc Lucer0 on 12.28.12 at 4:55 pm

Matt, would you happen to have a list of all of these databases that is not in a grpahic. I am interested in doing research on these DBs but I feel I might miss some if I just copy from the sheet as I am sure I’ll get some of these things wrong.

#13 Marilyn Matz on 12.31.12 at 10:31 am

Superb map, Matthew! Thanks for updating it.

As an array database, Paradigm4’s SciDB is a non-relational analytical database that should be pinned near the Berkeley Data Analytics Stack.

#14 Matthew Aslett on 01.02.13 at 5:00 am

Thanks for the clarification.

#15 Sam on 01.03.13 at 4:15 am

No SQL Server Analysis Services? That comes in two flavours now with the addition of tabular mode

#16 Michael on 01.03.13 at 6:56 am

Hi – I only know MS Access so far. Is it somehow a frontend to a database which can be found in the chart?

#17 Matthew Aslett on 01.03.13 at 9:10 am

Hi Michael,

Yes MS-Access can be used as a front end to MS SQL Server, as well as others such as Oracle and Sybase.

#18 Alex on 01.03.13 at 8:37 am

Great job! .. plz provide SVG!

#19 C4 on 01.03.13 at 9:05 am

Software AG’s ADABAS?

#20 Matthew Aslett on 01.03.13 at 9:10 am

Thanks C4

#21 Omer on 01.05.13 at 1:34 pm

To add to the “Grid/Cache Zone”

Amazon ElastiCache: It provides a managed caching in AWS and supports Memcached.

#22 Matthew Aslett on 01.06.13 at 5:24 pm

Thanks Omer,

I could have sworn I’d put that on, but clearly I didn’t


#23 Jerzy on 01.08.13 at 7:39 am

What does it mean “operational” and “analytic”? Isn’t CouchDB or Oracle NoSQL operational?

#24 Matthew Aslett on 01.08.13 at 9:40 am

Yes, indeed, they should be thought of as extensions to the traditional ‘operational line’. This is something that could be clearer

#25 Jerzy on 01.08.13 at 12:42 pm

If they are operational, so why aren’t they on the “operational” navy line on your map? What is the difference between extensions and core operational lines?

#26 Matthew Aslett on 01.08.13 at 12:49 pm

Hi Jerzy,

Because then almost all of the map would be the same colour and I wanted to make a distinction for each of the NoSQL/NewSQL branches etc

#27 Ostap on 01.10.13 at 5:56 am


#28 Michael Waclawiczek on 01.11.13 at 9:04 pm

Matt Aslett will be one of our key note speakers at NuoDB’s live product launch event on Tuesday, Jan 15, 11am EST. Please join us if you can –

#29 Jane on 01.13.13 at 9:46 am


#30 Michael Hummel on 01.13.13 at 6:58 pm

Thanks for adding ParStream to your map – well placed there as an analytical database providing ultra-high speed, full flexibility and continuous high speed data import.

#31 Michael Hummel on 01.14.13 at 2:25 am

Just saw that our company name is not spelled correctly. Maybe you can correct in the next version to ParStream.

ParSteam is actually a good name we could use for our cloud offering 😉 LOL

#32 Matthew Aslett on 01.14.13 at 4:58 am

Thanks Michael, and apologies. I’ll change it on the master.


#33 Priya Bhatt on 01.18.13 at 2:51 pm

Thanks Matthew,

Since this is a frequently read page, it would be great if the map with the right spelling of our name can be shown on this webpage as well, and also in your blog that refers to this map (“Exclusive guide to…”).
Much appreciated!
-Priya, AVP- Marketing