SQL Stress Tool Benchmark

Our first benchmark was custom-written in .NET, using ADO.NET to connect to the database. The AnandTech Forums database, which is over 14GB in size at the time of the benchmark, was used as the source database. We'll dub this benchmark tool "SQL Stress Tool" for the purposes of discussing what it does. We have done some updates to the tool since we first used it; it now supports Oracle, and MySQL. We also adjusted the test time for this test and future tests to 20 minutes. The reason for this was to ensure that we used as much memory as possible for future planned 64 bit tests.


Click to enlarge.

SQL Stress allows us to specify the following: an XML based workload file for the test, how long the test should run, and how many threads it should use in which to load the database. The XML workload file contains queries that we want executed against the database, and some random ID generator queries that populate a memory resident array with ID's to be used in conjunction with our workload queries. The purpose of using random ID's is to keep the test as real-world as possible by selecting random data. This test should give us a lot of room for growth, as the workload can be whatever we want in future tests.

Example workload:


    Select1
    select count(iuserid) as usercount from ftdb_forumusers where iforumid = 1


    Select2
    select count(u.iuserid) as currusercount from ftdb_users u,ftdb_forumusers fu where fu.iforumid = 1 and u.iuserid = fu.iuserid and dtlastvisiteddate > '[q]qGetLastVisitDate[/q]'


Example Random ID Generator:


    qGetLastVisitDate
    select dtlastvisiteddate,newid() as ldate from ftdb_users where dtlastvisiteddate is not null order by ldate


The workload used for the test was based on every day use of the Forums, which are running FuseTalk. We took the most popular queries and put them in the workload. Functions, such as reading threads and messages, getting user information, inserting threads and messages, and reading private messages, were in the spotlight. Each reiteration of the test was run for 20 minutes, with the first being from a cold boot. SQL was restarted in-between each test that was run consecutively.

The importance of this test is that it is as real world as you can get; for us, the performance in this test directly influences what upgrade decisions we make for our own IT infrastructure.

Web Tests - FuseTalk .NET SQL Stress Results
Comments Locked

144 Comments

View All Comments

  • Phlargo - Thursday, April 21, 2005 - link

    Sounds like computing could change a lot in the next 6 months.

    I can't wait to get my hands on a dual core chip - that type of multitasking gaming performance is what I've been waiting for... 92% of unfettered performance during Doom3? That's unbelievable.
  • manno - Thursday, April 21, 2005 - link

    Another thing to consider is that the 939, and 940 Dual Cores work with old mobos. so that saves upgraders ~$100 for the mobo (I have no clue what PD mobo's are going to cost) plus the cost of DDR2 ~$100 for 512MB, not true if you're going with the PD. So if you own a 939 a 2.8 PD will realy cost you ~$441, 3.0 -$516, 3.2 ~$730.
    Throw that out the window if you're starting from scratch however.
  • manno - Thursday, April 21, 2005 - link

    Great review very detailed, I have one caveat, and I should preface it with the fact that I own 4 systems, and all of them have Athlon 64's or XP's.

    That being said your "Multitasking Scenario 2: File Compression" Seems to be misleading, and identifies the A64 as the processor of choice in the 2nd portion of the analysis.

    Maybe I'm wrong but from what I understand with how the test was performed you archived the file, got the amount of time it took to archive the file "x" for instance, and then figured out how many emails got imported in time "x" now because some processors took longer to create the archive than others they had more time to import emails. so in order to make this data more reflective of the performance of each processor you need to divide the number of emails imported by "x", and get e/s. The modified graph should be

    PD 3.2 =19124/5.08 or 3764.57 e/s
    A64 X2 =21687/6.25 or 3469.92 e/s
    PEE 3.2 =16875/6.65 or 2537.59 e/s
    AFX 55 = 3800/5.88 or 646.26 e/s

    These corrected numbers show that at least in this test the PD 3.2 is the winner by ~8%

    -manno
  • Calin - Thursday, April 21, 2005 - link

    At that prices, most of the users really won't need their current mainboards for new dual core processors. I would prefer to have lower priced single core than that prices on dual core.
  • blckgrffn - Thursday, April 21, 2005 - link

    We are talking about by this fall, folks. And a 90nm Dual core still uses LESS die than 2 2800+ put together because it is on a manufacturing smaller process, especially if they stick to a 512k cache which is obviously the sweet spot. By this fall I expect that we will see sub $200 dual cores from Intel to squeeze AMD on the desktop market. What is a real shame is that AMD doesn't have suffcient fabs to handle potential demand.

    For those of you who got all hot in your pants and said I can't buy a sub $200 dual core Intel right now, it is just as true that I can't buy one it all! ;P Chill!
  • PeteRoy - Thursday, April 21, 2005 - link

    Impressive.
  • josedawg - Thursday, April 21, 2005 - link

    Am I mistaken in thinking that AMD was throwing out their desktop dualcore line (A64 X2) in 2006? What prompted their earlier release of them? Was it the demand for desktop dualcore, pressure from Intel stealing desktop dualcore market, or excellent manufacturing at the AMD plants?
  • KeithDust2000 - Thursday, April 21, 2005 - link

    blckgrffn, "so if they can't bring anything out under $200 I will probably have to go with Intel. Boo for that ;) "

    INTEL doesn´t have one under $200 either.
  • Jeff7181 - Thursday, April 21, 2005 - link

    #14... all socket 939 motherboards will support the X2 with a BIOS upgdate.
  • Zebo - Thursday, April 21, 2005 - link

    #9 They'd have to release a 1.6 @ $240 to accurately compete with Intels slowness in Pentium D's starting at $241. Sorry not gonna happen. AMD not for budget shoppers anymore but those interested in performance. Want the best? Pay the price. Or substandard CPU at a discount?

    I can see a 1.8 for about $250 though however no way you're going to get into DC for less than $200.

Log in

Don't have an account? Sign up now