[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference ulysse::rdb_vms_competition

Title:DEC Rdb against the World
Moderator:HERON::GODFRIND
Created:Fri Jun 12 1987
Last Modified:Thu Feb 23 1995
Last Successful Update:Fri Jun 06 1997
Number of topics:1348
Total number of notes:5438

664.0. "Oracle ten times faster on Sequent than VAX (PR)" by IJSAPL::WOODROW (Father of DRS) Sun Jun 03 1990 17:53

Headline in the May 23rd edition of the Dutch version of Computerworld:

   Performance Oracle on Sequent ten times better than on VAX

(my translation)

The article describes a large Dutch company which had been working with
Oracle on a 3500 and was not satisfied with performance. One choice was
a VAX 6300 at 1 million gulden, but Oracle provided major assistance in
benchmarking a Sequent (AND Oracle 6 with row locking!) at 750 thousand
gulden. An SQL report for invoices that required 20-30 minutes on the
3500 was ready in 3 minutes on the Sequent. Sequent is in.

The same article makes a point of mentioning that the company also uses
an IBM 4331 - which is still running satisfactorily!

Wether this article was the cause or not, we now have a major world-wide
strategic account that has put a freeze on all expansion of Digital
products until it is assured that the upgrading process will not go on
indefinitely. Furthermore, Oracle is doing a benchmark for them on
Sequent next week!

Oracle does it again.
T.RTitleUserPersonal
Name
DateLines
664.1Sequent IS faster than 3500 !!MAIL::DUNCANGOracle... the one-line databaseSun Jun 03 1990 23:3323
    Well what did you expect ?  uVAX 3500 is about 2.4 mips and the
    smallest Sequent is, I believe, 7 mips but it is common to use the 18
    mip Sequent for tests.  So when you take the wide MIP rating into
    account and factor in the fact that qbus i/o is not very good, it's
    all perfectly logical.
    
    I can tell you from first hand experience that a 6000-420 will beat
    an 18 mip sequent EASILY and that a 6000-440 will be the 24 mip Sequent
    EASILY.  The problem is that we win the performance war and lose the
    price/performance war due to the Oracle license fee game.  Based on
    my customer's batch benchmark with Oracle V6/TPO on both the VAX
    and the Sequent, our 6000-4xx won EVERY time .... and
    
    ... due to our growth path, we won the order !
    
    One other problem you should be aware of is that Oracle receives 
    fees from Sequent, DG, HP, Pyramid, and many others when their database 
    software results n hardware sale.  Digital is about the ONLY hardware
    company that doesn't have this plan for Oracle.  You tell me who
    Oracle is going to help and why there is such a wide delta in license
    fees ????
    
    -- gerry
664.2So what about DEC U* vs Sequent U* ??37742::POWELLReed B Powell 422-7291 PTO Sales SupportThu Jun 07 1990 19:225
    Any thoughts out there on how good the 5000/200 would be at beating
    SEQUENT on the U* platforms?  Seems to me that we have a good story
    even on the U* platforms in terms of growth:  Even with 486 technology
    in the Sequent line, we can push that window further and quicker with
    future  RISC products.