IBM Informix - SQL & NoSQL
Matching: hana X
Performance and scalability of Informix ultimate warehouse edtion on Intel Xeon 7500 and E7 processors
keshavamurthy 0600019DXM Tags:  warehouse exalytics oracle intel accelerator scalability hana informix xeon sap performance 3,673 Views
keshavamurthy 0600019DXM Tags:  ibm informix sap accelerator ex5 warehouse hana 2 Comments 4,960 Views
The Doloitte technology trends 2011 report point to new in-memory technologies like SAP HANA changing the business intelligence is done.
This trend bodes well for IBM Informix warehouse accelerator (IWA) we recently released. SAP HANA and IWA both try to exploit in-memory systems to accelerate query performance and reduce administration+tuning at the same time. These used to be inversely correlated, but with 3rd generation database technology, we're trying to get both together! It's like talking and browsing together on the phone! :-)
I saw one of their videos indicating queries on 700m row fact tables returned in 5-7 seconds... IWA has similar numbers with queries on more than billion row fact tables returning 2-4 seconds. IWA is transparent to Informix applications, runs in virtualized environment, on the cloud, etc -- are plus points for IWA. It'll be interesting to see if someone spends time to do a full blown comparison.
The improvements in hardware configurations are aiding this trent as well.
Recently, IBM announced new versions of eX5 systems. New MAX5 expansion capabilities include:
• Eight-socket x3850 X5: up to 192 DIMM slots (6 TB)
• Four-socket x3850 X5: up to 96 DIMM slots (3 TB)
• Two-socket x3690 X5: up to 64 DIMM slots (2 TB)
• Two-socket HX5: up to 32 DIMM slots (512 GB)
Having these large memory systems enable more VMs and enable improved CPU usage.
These systems bring down the overall cost and improve the cpu utilization for existing software.
New generation of software like IWA automatically maximize the CPU usage and larger memory systems enable larger data warehouse.
The eX5 tagline is:Maximize your memory, minimize cost, simplify deployment.
This -- and maximize performance -- could very well have been tag line for IWA as well :-)