1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

Test Signal Processing Power of your PC

Discussion in 'Recording' started by audiohead, Apr 15, 2002.

  1. audiohead

    audiohead Guest

    Vincent Burel wrote a simple proggie to test your PC capabilities in real audio processing situations. Test results look a bit abstract to me, but if we post them here we can compare how our DAWs score. I'm particularly interested in ANUS DAW scores (PIV 1.6 FSB versus o/c).
    Get this proggie at http://www.vb-audio.com and let's start the race :)
     
  2. Opus2000

    Opus2000 Well-Known Member

    Hmmm..interesting...will definately check it out. I just redid my system on Win2k last night..wish you had posted this before I switched from Xp so I could see a difference between the two OS'. I'll do this when I get home from work and let you know
    Opus
     
  3. Nick Driver

    Nick Driver Guest

    Oh goodie! let's see who can pee the farthest!

    System: Rambus ANUS (AsusNorthwoodUltraSystem)
    ASUS P4T-E, 512MB Samsung RDRAM, 1.6a, OC'ed to 2.128GHz (FSB=133MHz, mem=x3=399MHz)
    ATI Radeon 7500 64MB, oem version
    Windows XP Pro

    Test SPP benchmark results:

    StreamType SPP-1 measure
    ------------- ---------------
    mono 44.1k 1263 cycles/smp
    mono 48.0k 1160 cycles/smp

    StreamType SPP-2 measure
    ------------- ---------------
    mono 44.1k 1505 cycles/smp
    mono 48.0k 1383 cycles/smp

    StreamType SPP-M measure
    ------------- ---------------
    mono 44.1k 6426 cycles/smp
    mono 48.0k 5904 cycles/smp
    Transfer Rate 540 Mfloat/s

    I don't really feel like taking the cover off to reset the dipswitches to test the normal 1.6GHz speed right now, sorry.
     
  4. audiohead

    audiohead Guest

    OK, this is the very basic ANUS setup: P4 1.6a, ASUS P4266-SE, DDR Samsung 256 Mb(pc2100) @CL=2.5, ATI 3D Radeon VE, nothing

    Test SPP benchmark results:

    SPP-1
    ------------- ---------------
    mono 44.1k 903 cycles/smp

    SPP-2
    ------------- ---------------
    mono 44.1k 1069 cycles/smp

    SPP-M
    ------------- ---------------
    mono 44.1k 4663 cycles/smp

    Transfer Rate 392 Mfloat/s

    At first I thought I peed quite far,
    but now that Nick took out his d***

    (oops, I better shut up :D )
     
  5. Opus2000

    Opus2000 Well-Known Member

    Well...here's the tests with XP...
    Tested with OS on Primary Controller(WinXP PRO ACPI Mode)

    Stream Type SPP-1 Measure

    Mono 44.1Hz 1235 Cycles/smp.
    Mono 48 Hz 1135 Cycles/smp.
    _________________________________________

    Stream Type SPP-2 Measure

    Mono 44.1 Hz 1461 Cycles/smp
    Mono 48 Hz 1343 Cycles/smp.
    __________________________________________

    Stream Type SPP-3 Measure

    Mono 44.1 Hz 5890 Tracks
    Mono 48 Hz 5411 Tracks

    Transfer Rate 495 Mfloat/s

    And then.....

    Tested with OS drive on RAID(WinXP PRO ACPI Mode)

    Stream Type SPP-1 Measure

    Mono 44.1Hz 1238 Cycles/smp.
    Mono 48 Hz 117 Cycles/smp.
    _________________________________________

    Stream Type SPP-2 Measure

    Mono 44.1 Hz 1455 Cycles/smp
    Mono 48 Hz 1336 Cycles/smp.
    __________________________________________

    Stream Type SPP-3 Measure

    Mono 44.1 Hz 5891 Tracks
    Mono 48 Hz 5412 Tracks

    Transfer Rate 495 Mfloat/s

    As you see..booting off of the RAID controller only gives you a slight improvement..also bear in mind that the OS drive is on the primary controller on the mainboard by itself for the first test...
    I was going to do 2k tests but my system has to go out on another short rental! :(
    Oh well...mo moeny mo money mo money!! lol
    Opus
     
  6. Tommy P.

    Tommy P. Well-Known Member

    XP Pro ACPI mode, 512Mb Crucial set to cas2.

    Stream Type SPP-1 Measure

    Mono 44.1Hz 1236 Cycles/smp.
    Mono 48 Hz 1136 Cycles/smp.
    _________________________________________

    Stream Type SPP-2 Measure

    Mono 44.1 Hz 1460 Cycles/smp
    Mono 48 Hz 1342 Cycles/smp.
    __________________________________________

    Stream Type SPP-3 Measure

    Mono 44.1 Hz 5907 Tracks
    Mono 48 Hz 5427 Tracks

    Transfer Rate 496 Mfloat/s

    The most important criteria for me in a DAW over processing power is compatability, and glitch free real world use. LOL, thats why I went from a kick ass AMD system to a kick ass SDFC system.
    All the benchmarking going on in mainstream hardware review sites, has lost track of what the use of the machine is in the first place.
    This proggie by Vincent Burel, is useful, but getting a little old. WTF, RAID or cas2 doesn't budge the meter? LOL :p

    Tommy P.
     
  7. Dedric

    Dedric Guest

    FWIW,

    System: Athlon XP 1900+
    Gigabyte GA7-DXR+ board
    1G Muskin PC2100 SDRAM
    Windows XP Pro

    Test SPP benchmark results:

    StreamType SPP-1 measure
    ------------- ---------------
    mono 44.1k 1511 cycles/smp
    mono 48.0k 1389 cycles/smp

    StreamType SPP-2 measure
    ------------- ---------------
    mono 44.1k 1453 cycles/smp
    mono 48.0k 1335 cycles/smp

    StreamType SPP-M measure
    ------------- ---------------
    mono 44.1k 2639 cycles/smp
    mono 48.0k 2424 cycles/smp
    Transfer Rate 222 Mfloat/s

    Seems comparable in SPP-1 tests for the most part, but significantly lower in SPP-M. I don't know if that means my Athlon will stream 1/2 the tracks of a Northwood or not, but there it is. However, if this test indicates system problems it could be helpful - i.e., should the SPP-M test have higher numbers for an Athlon XP 1900? (Seems like it is more raw processing power, which should be less configuration dependant than an application's work would be).

    I'm not sure it matters as I've streamed 48 continuous stereo tracks with plugins, no problem, which is more than most mixes will have active at one time. Stability hasn't been an issue (other than a few Logic 5 related crashes under specific conditions), and latency isn't either with an HDSP card.

    Dedric
     
  8. Tommy P.

    Tommy P. Well-Known Member

    Dedric, this benchmarking program is already one year old. It may not be optimized, hell,it could be faulty in any number of areas. At best, its useful for comparing similar machines, to raise a flag, or point out a config problem. :cool:

    Look at all the releases of Sandra, and how the results can vary from one version to another.
    Just my 02.
     

Share This Page