Im architecture: CPU architecture: SIMD vector processing (im-2.3)

Continued from Chapter 2 Oracle Database In-Memory Architecture (IM-2.2)

This article is IM Architecture Part II

CPU framework ：SIMD Vector processing （Vector Processing）

For the need to be in IM Data scanned in column store , Database usage SIMD（ Single command , More data ） Vector processing .

IM Column storage maximizes the number of column entries that can be loaded into vector registers and evaluated . Instead of evaluating each entry in the column one at a time , The database is in a single database CPU Evaluate a set of column values in an instruction . SIMD Vector processing enables the database to scan billions of rows per second .

for example , The application issues a query to find sales Used in table promo_id The value is 9999 Total number of orders .sales The watch resides in IM Column storage . Query by scanning only sales.promo_id Column start , As shown in the figure below ：

chart 2-12 SIMD Vector processing

CPU Calculate the data as follows ：

1. Before the 8 It's worth （ The value varies according to the data type and compression mode ） from promo_id Column loading SIMD register , Then compare them with the values in a single instruction 9999 Compare .
2. Discard entries .
3. In addition 8 Values loaded to SIMD In the register , Then continue in this way , Until it has evaluated all entries .

Shandong Oracle User group (Shandong Oracle User Group), abbreviation ：SDOUG, Is a full of vitality 、 Young non-profit organizations , It aims to provide an exchange platform for technology lovers in Jinan and surrounding areas .SDOUG Organize offline technology sharing activities from time to time , Promote local and surrounding IT Technological development 、 Help technology enthusiasts improve themselves . Share technology 、 Share happiness ,SDOUG On the road .