• Register
  • Help
Results 1 to 6 of 6

Topic: WD 38GB raptor VS WD 40GB 8M cache ?

  1. #1

    WD 38GB raptor VS WD 40GB 8M cache ?

    hmmm, I am running WD 40GB-JB (8M cache version) , I am wondering if I could gain a big benfit from raptor series HD, does anyone could give me come comment please ? ( RPM mean nothing tho )

    I use it for cubase only, Thanks.

  2. #2

    Re: WD 38GB raptor VS WD 40GB 8M cache ?

    Hi, the exact amount of performance boost (track,voice count etc) is hard to predict.
    Anyway 10000rpm of the raptor vs 7200rpm of the WD40means about 30% faster disk access times.
    The problem is that disk access time is not the only parameter that determines the global performance of your HD recording app or disk streaming sampler.
    The other parameter is sustained disk throughput alias how many MB/sec the disk can transfer.
    This parameter depends not only from the disk but from the speed of the bus that transfers the data between disk and CPU (IDE, SCSI etc) plus by the efficiency of the operating system in transfering the data (file cache, etc) and finally the application itself.

    If you have a disk with not so good access times but decent throughput you can compensate for this by using large RAM buffers when streaming from/to the disk, provided the application lets you adjust these settings.

    BTW: speaking of disk performance measuring.
    We wrote some benchmarking code which we use to profile and optimize LinuxSampler, I toyed with the idea to enhance it a bit and release an easy to use free, crossplatform (works on Windows, Mac OS X and Linux) disk benchmark which measures the performance of a disk simulating the access patterns generated by HD recording apps and disk streaming samplers.

    The benchmark currently tells you what's the real world average disk seek time, the data throughput in MB/sec by streaming multiple tracks from disk, how many tracks you can stream, what's the optimal buffersize and how much RAM is required to stream the maximum achievable simultaneous tracks.

    I searched the net for such benchmarks but I did not find anything useful.
    The advantage of my cross platform benchmark is that you can compare results on different operating systems, you could before you buy a certain hardware run the benchmark and see what kind of numbers you get out of it and it would make it easier to decide if a PC or a Mac gives you more performance per dollar/euro

    Then an additional step would be let users submit their results so that we have a nice database of systems and their performance (eg systems with RAID arrays, fast SCSI controllers etc).
    That way it becomes easier for users to make buying decisions.

    Anyone interested, additional suggestions ?

    PS: we got a preliminary GUI


  3. #3

    Re: WD 38GB raptor VS WD 40GB 8M cache ?

    FWIW, here's a recent Raptor review...

  4. #4

    Re: WD 38GB raptor VS WD 40GB 8M cache ?

    quote from the article:
    "Our tests have also shown us that the 10,000RPM Raptor can offer a noticeable, but not dramatic, performance improvement over the current generation 7200RPM 8MB cache drives. While the performance improvement is there, it's not as significant as the synthetic tests would have you believe."

    Perhaps this is true in general but specific applications that perform lots of disk seekings (HDR apps, disk based samplers) will certainly benefit more than a game loading large level files in a linear fashion.

    I think I'll go ahead with implementing my benchmark, which at a later stage could be proposed to sites like anandtech.com, tomshardware.com for inclusion in their standard benchmark suite.


  5. #5

    Re: WD 38GB raptor VS WD 40GB 8M cache ?

    Quote Originally Posted by sbenno
    I think I'll go ahead with implementing my benchmark, which at a later stage could be proposed to sites like anandtech.com, tomshardware.com for inclusion in their standard benchmark suite.
    Excellent. In the meantime I e-mailed them about GS3 Orchestra as a benchmark. It's an easy test to load up some bundled sounds, run an insanely busy MIDI file, then check the peak poly. As long as they use the same libs loaded the same way on defragged disks and the same MIDI file in the same system, the disk test should yield a valid number.

    I'd bet that your benchmark test would be cheaper though.

    In any case it will be awesome to get real world sampling benchmarks into the hands of the disk testers.

  6. #6

    Re: WD 38GB raptor VS WD 40GB 8M cache ?

    While GS3 would be a real world benchmark I think those hardware testing sites would probably not too much willing to use it, both because they are probably not audio/sampler experts and because they want automated benchmarks not guessing and reading poly peaks from a running app.

    Plus softsamplers perform DSP work too which is CPU dependent so the total poly is a combination of both disk and CPU speed which makes it harder to isolate the pure disk speed advantage.

    PS: just for curiousity how many voices (GSt or Kontakt, specify if the sample used is mono or stereo) do you get out of a modern laptop with internal ATA100 drive ?
    on My Acer 1500 (Athlon 64, 60GB HD) my disk streaming benchmark gives me about 80-90 stereo voices max. (in GSt terms 160-180 voices)


Go Back to forum


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts