There’s a lot of activity surrounding enterprise analytics and ‘big data’ these days. The major IT vendors like IBM, HP, and Oracle have all bulked up their software, system, and service offerings with an eye toward being a one-stop enterprise analytics shop. Cisco and Dell (and others) will probably start making some more noise …
Another emperors new clothes story.
As for other IT sectors there is very little that's truly new in the market. Big Data is just the Web 2.0 fad relabelled for analytics.
Teradata are doing nothing new, and only play in the foundations for Analytics - they are a Niche Datawarehousing player that's all. A very good niche player who can justify the premiums they charge but Niche nonetheless.
Even IBM and Oracle are doing nothing new. Both are struggling to merge their acquisitions into coherent stacks on the sales, functional and technical levels.
In summary - a bit more consolidation in the Analytixs space, and revolutionary Big Data use cases are very few and far between - even for most corporates. Big Data does not simply mean a big DWH.
SAP Sybase IQ
Surprised you didn't mention Sybase IQ. Also designed from the ground up to be a data warehouse. It seems to be growing quite fast and seems a pretty good product.
Being picky, according to their own timeline, Teradata wasn't acquired by NCR until 1991: http://www.teradata.com/history/.
The NCR years can be 'blamed' for Teradata being wedded to NCR's servers and 32bit Unix MP-RAS for too long. The relatively recent move to OEMd Dell servers running 64bit SUSE Linux with lots of RAM was very welcome.
In addition to the ability to support a "large number of nodes" - tightly coupled to support the scale-out/MPP architecture - Teradata's other differentiators are intra-node parallelism through the use of virtual processors and the 'bynet' interconnect to enable high speed, scalable and reliable inter-node data movement. See http://it.toolbox.com/wiki/index.php/BYNET.
As the author says, Teradata's "tight focus on data warehousing and data analysis" and the fact that the DBMS is "designed from the ground up for data warehousing and decision support" are clear differentiators.
General purpose databases used for decision support a) generally don't scale out and b) were never designed to support high-speed, high-volume join, aggregate, sort and scan operations. Add complex/ad hoc queries and high concurrency to the mix and general purpose databases soon struggle.
IBM with Netezza, EMC with Greenplum, Microsoft with Datallegro/PDW and Oracle with Exadata are clearly aiming for a slice of the descision support action. After ditching Neoview, HP bought Vertica so they are still theoretically in the race.