Performance Across All Tracks

Hard drives are made up of one or more circular platters. Platters are written from the outside inward in order to maximize performance (you cover more data in a single rotation of an outer track vs an inner track). I used HDTach to characterize the new VelociRaptor's performance across all LBAs:

Average read/write speed over the drive's capacity is around 130MB/s. The minimum sequential speed you'll see is around 90MB/s, while the maximum is over 140MB/s. The burst speed here is only 213.6 MB/s, which isn't enough to saturate 3Gbps SATA. In my tests, using the X58's 3Gbps controller yielded better performance with the new VelociRaptor than hooking it up to a 3rd party Marvell 6Gbps controller.

While hard drives are starting to embrace the 6Gbps standard, it's simply not necessary from a performance standpoint.

Random Read/Write Speed Overall System Performance using PCMark Vantage
Comments Locked

77 Comments

View All Comments

  • pjconoso - Thursday, April 8, 2010 - link

    Maxtor's with Seagate, so there.
  • ClagMaster - Tuesday, April 6, 2010 - link

    These are good drives carrying on the Raptor tradition, but at the asking price at ~$0.60/Gb, way too pricy.

    I am going to wait next year for third generation SSD's at $0.50/Gb and 2-4x the current bandwidth before switching over.
  • leonsk - Tuesday, April 6, 2010 - link

    Agree with Anand that SSD bor boot drive and 2 large HDs in RAID-1 is the most effective rig. What controller do you use? Or are you with software RAID?
  • AstroGuardian - Wednesday, April 7, 2010 - link

    Of course not mate. Software RAID is dead. I am sure he is using ICH10R's RAID controller.... Right Anand?
  • bobbyto34 - Wednesday, April 7, 2010 - link

    Hello,
    I have to change my primary hard drive (Samsung Spinpoint F1) because of some read failures...
    I'm hesitating between a good ol' 1TB 7200 drive, an x-25M 160Gb or a new Velociraptor.

    The computer will be used as music station with Sonar and sound banks. The current problem is that it's quite slow when loading those enormous sound banks with a 7200 drive.

    I would like good performance in loading these sound banks, with silence because it's dedicated to music (and some gaming :p), and with reasonable disk space... For the moment, I can't see a good compromise with those 3 parameters. Am I wrong ?
  • slickr - Wednesday, April 7, 2010 - link

    What the heck were you talking there buddy?
    You just buy 2 1TB HD and an SDD drive, to be faster. Do you know how much such configuration would put you back?

    If not, go for the new raptor. Oh really, you just found out hot water. How about you mention its twice as cheap and even more to go for it, ten times more reliable than SSD's, and you can actually store 600gb, instead of 80gb of the intel ssd's or whatever for 250 dollars.
  • Compddd - Wednesday, April 7, 2010 - link

    When will these start shipping?
  • GullLars - Wednesday, April 7, 2010 - link

    Another great test from Anand.
    I have a comment though, and i see i'm a day late, so i may be ignored here :(

    The IOmeter sequential tests, while spanning the entire drive, only run for 3 minutes, and therefore only uses a small % of the drive area, so the speeds you see are near max.
    For the 600GB velociraptor, you have 140MB/s for 3 minutes = 25GB. By looking at the HDtach curve, the drop-off has barely started by that point.

    I also have something i want to get out there. Off-topic to this test, but relevant for high-performance HDDs.
    Why hasn't any HDD maker yet added a flash read-cache?
    By adding a single 4-8GB MLC NAND chip, costing roughly $2-3/GB = $8-24 added cost, and using it for read-caching hot-files, you can get around 4-5000 4KB random read IOPS = 16-20MB/s (@QD 1) and roughly 40-60MB/s sequential read for the cached data.
    Tracking hot-files should be easy to implement simply by logging read-access to LBAs, and with a slight bit more effort, filtering LBAs being read in a small block random pattern. Possibly also caching file-table and folder/file structure and metadata, as well as the data typically read the first seconds after power-up or spin-up.
    Using this type of caching would have a noticable effect on typical usage patterns, and especially multi-tasking, but would likely not make an impact on benchmarking since it would take time for new data to reach cache.

    Larger hot-files could benefit from the cache as they could be read from both flash and disk at the same time with the speed of both combined (80-140 + 40-60).

    Any thoughts about using flash read-cache people?
  • ETR - Wednesday, April 7, 2010 - link

    Thats all i want to know,

    Does this drive have it or not???

    not one reviewer anywhere has that info... nor did WD's website.

    gezzzz how can anandtech skip that info................?
  • BoFox - Thursday, April 8, 2010 - link

    Dear Anand,

    I'm looking to upgrade to 15000rpm SAS drives since my Foxconn Bloodrage motherboard has 2 SAS ports (it's an X58 motherboard and there are couple other from Asus and Gigabyte that include SAS ports, IIRC..).

    How would this 600GB Velociraptor fare against a Seagate 15k.7 600GB Cheetah in terms of speed/performance?

    Sure, a 15k drive might be a bit more noisy and power-hungry, but I'm ok with this since I have other noisy stuff (including a noisy blu-ray player). I'm thinking that having a couple 15000rpm 600GB SAS drives is still a better investment than those tiny SSD's. Over at techreport.com, it's shown that SSD's (including Intel's) have some serious performance issues with certain real-world applications and that they "degenerate" in performance whether TRIM is being used or not. Stability is another question, right?

Log in

Don't have an account? Sign up now