Some of you may be surprised to read I've recently become less religious about Teradata. That's not because Teradata has declined or it's competitors have improved significantly.
It's mainly because I see the market moving toward realtime and near realtime reporting. That drives to enhancing OLTP systems for performance to allow them to do OLAP on or near the system of record (think 64 bit addressable memory and solid state drive technology) .
After 20 years of RDBMS performance that didn't meet expectations, I'm predicting the beginning of the end of Data Warehousing and those evil replicating Data Marts. As the performance curves are finally catching up to the business' BI expectations. Data Warehousing as we know it will gradually fade away. A lot of people in DW and BI aren't ready to hear that from me… yet. Just remember you heard it here first.
The focus now shifts to the Enterprise Architecture, Enterprise Metadata and Systems Integration Architectures, good-old Data lineage and that old favorite Referential Integrity on an enterprise level. Woow!
I am wondering what big change you see that will make all this possible. It certainly would change Business Intelligence.
@mip69 re cloud: network bandwidth is the constraining factor in big data across the cloud to build the “virtual DW”. all we need is lots of dark fibre.
That is an exciting trend. Do you think that cloud-based services will also play a role in the elimination of the classic EDW? Great post!