Updated July 2013: See here, here and here for details.
| Vendor/Product |
Columnar Maturity |
| Teradata Database |
2 |
| Oracle Exadata |
1 |
| SAP HANA |
3 |
| Pivotal Greenplum/HAWQ |
2 |
| IBM DB2 BLU |
3 |
| Microsoft SQL Server xVelocity |
2 |
| HP Vertica |
3 |
| Actian Paraccel |
3 |
| IBM Netezza |
n/a |
| SAP Sybase IQ |
3 |
| Infobright |
1 |
| Vectorwise |
1+ |
Level 1 Columnar: Uses PAX to achieve columnar compression. No columnar projection provided. No columnar engine provided. Approximate 4X performance advantage over row store for read queries (10X column compression versus 2.5X row compression).
Level 2 Columnar: Uses columnar compression and projection. No columnar engine provided. Approximate 10X advantage over Level 1 read queries (10% of the columns are selected).
Level 3 Columnar: Uses columnar compression and projection… and includes a columnar engine that optimizes processing. Approximate 50X advantage over Level 2 read queries (Vector processing – 20X, SIMD – 8X, Fewer CPU Stalls – 2X, Cache Utilization – 10X, in-memory compression + projection 20X in differing combinations for each query)
Oracle says that only the relevant columns are projected using Smart Scan. However, here you are saying that all the columns are read when a table is queried. Can you clarify the ambiguity.
http://docs.oracle.com/cd/E24693_01/doc.11203/e15883/exadata.htm
Hi Omair,
Smart Scan in an Exadata feature that projects from the Storage layer to the RAC layer the columns as you suggest. The 12c In-memory feature is not an Exadata feature. It resides in a regular Oracle or in an Oracle RAC configuration… and the columns stored in-memory are statically selected by the DBA.
Rob