Home Page for the TeradataForum
 

Archives of the TeradataForum

Message Posted: Fri, 13 Jun 2003 @ 14:53:18 GMT


     
  <Prev Next>   <<First <Prev Next> Last>>  


Subj:   Re: Netezza vs Teradata
 
From:   McBride, Michael

Doesn't it make sense that what Netezza has done is revert to Teradata's (not NCR's) original concept, that is to have a "database machine" (remember the DBC 1012), using hardware (embedded software) to perform the low level task of managing data, rather than software (traditional RDBMS). They are going back to a great idea, and building upon new technology (faster, smarter cpus's, bus speeds, memory speeds, bigger faster disks, disk controllers, the whole hardware "inside" the box has advanced so far over the past 20 years or so since Teradata first introduced AMPs and the Ynet.

When NCR (after acquiring Teradata) decided to migrate from a "closed" proprietary architecture, to more "open" software solution along the lines of DB2 and Oracle (software RDBMS on top of server technology) they effectively limited performance to the execution speed of software "code".

I know I'm probably stating the obvious, but it appears that Netezza is just following the logical path Teradata would have, had not someone decided to make the Teradata RDBMS a UNIX system software program rather than advancing and enhancing the AMP/COP/PE hardware/circuit board concept. Considering how fast hardware has advanced in just the last five years, this approach is like a reverse paradigm shift (going back to a great, original idea). I say this only because I perceive that the execution ability of hardware/firmware (embedded software) will perform much "faster" than executing compiled "c"code running on Unix or even Windows for that matter.

I'm certainly not a hardware expert, although I did start out on am 8 amp DBC 1012 in 1989, I know that even the first optimizer for Teradata was a program running under TOS (Teradata Operating System). But at the lower levels, actual data manipulation was more a hardware function than a software one (and the proprietary "intelligent" ynet (embedded software). If there is an original Teradaton out there (prior to the NCR acquisition), I would be interested in getting your thoughts? Am I way out of whack here, or am I somewhat close to the truth?

mem



     
  <Prev Next>   <<First <Prev Next> Last>>  
 
 
 
 
 
 
 
 
  
  Top Home Privacy Feedback  
 
 
Copyright for the TeradataForum (TDATA-L), Manta BlueSky    
Copyright 2016 - All Rights Reserved    
Last Modified: 27 Dec 2016