Although OCZ was first on the market with a USB 3.0 enabled SSD with its Enyo drive, competitors are knocking down the doors and bridging the gap. We saw the first lower cost USB 3.0 SSD with Kingston's DataTraveler Ultimate 3.0, however the JMicron JMF612 controller it was based on didn't really impress. The only thing the DataTraveler Ultimate 3.0 had going for it was a lower total cost for the smaller capacity versions.


The sole LED on the HyperX MAX 3.0 turns green when running in USB 2.0 mode, blue when in 3.0 mode


The drive comes with a USB 3.0 cable, don't lose it as there's no USB 2.0 port

Hot on the heels of the release of its unusually potent V+100 SSD, Kingston aims its sights at the high end with its second USB 3.0 SSD: the HyperX MAX 3.0. Borrowing a brand from Kingston's enthusiast memory line, the HyperX MAX 3.0 is literally a SSDNow V+100 SSD paired with a SATA to USB 3.0 bridge PCB:

The blue PCB is near identical to what we saw in the V+100 we reviewed:

Except in this case it plugs into a secondary PCB that bridges the 3Gbps SATA interface to a USB 3.0 interface:

The end result is an external SSDNow V+100 for those who want to carry around a decent amount of data and want very fast access to it.The HyperX MAX 3.0 uses the same Toshiba T6UG1XBG controller from the V+100. In our review of that SSD I found the new Toshiba controller actually did very well in real world workloads as well as maintaining sequential write speed even when peppered with random data. I originally attributed its behavior to a very aggressive garbage collection policy, however there may be other elements of Toshiba's architecture at work here (more on this in an upcoming article, hopefully).

The V+100 is probably the best platform you can think of today for an external SSD because its performance is so resilient. I took the HyperX MAX 3.0, filled it to its capacity, then wrote random data at a queue depth of 32 IOs across the entire drive for 20 minutes. While this process tanked random write speed on the drive, sequential performance was nearly untouched:

What this means is that under normal use as an external drive, it doesn't really matter what you copy to the drive or how often you use it. The drive won't lose sequential read or write performance - you'll still be able to copy movies, photos or whatever large files you've got at full speed. In fact, the worst I ever saw performance drop to was ~135MB/s after a bunch of random writing. After less than an hour of idle time the drive was back up to writing at full speed again across the entire drive.

Performance & Final Words
POST A COMMENT

30 Comments

View All Comments

  • name99 - Wednesday, November 24, 2010 - link

    "It does/should. 2.5" external HDDs need way more power to start spinning. This does work on most PCs."

    Ahh, the naivete of those who are convinced they carry an accurate mental model of the world in their heads.

    Go LOOK at the power utilization figures of a range of SSDs as against a range of 2.5" disks some time. You might be more than a little surprised at what you see.
    Reply
  • Chloiber - Thursday, November 25, 2010 - link

    Well, why don't you share your knowledge with me and post a link. Reply
  • name99 - Thursday, November 25, 2010 - link

    I would like to think that AnandTech users can use Google. Apparently not.

    Here's a table of SSD power consumption for a spread of devices. Note that at the peak end, we have utilizations of 3.7 to 4.1 W (for comparison a USB2 connector can deliver 2.5W, and FW 7W).
    http://www.anandtech.com/show/3812/the-ssd-diaries...

    A reasonable modern 2.5" magnetic drive, in contrast, draws 2.5W (basically engineered, I think, to run at USB2 power). Certainly our common experience is that Seagate and WD sell 2.5" drives that run just fine off a single USB connection.
    http://www.seagate.com/www/en-us/products/laptops/...
    Reply
  • blandead - Wednesday, November 24, 2010 - link

    can someone make an 8gb micro SDHC card this fast? That would be amazing, they work through USB as well once plugged in right? Or same sort of thing at least?

    anyways kudos to Kingston I hope prices go down faster I'm really itching to buy an SSD for my new build soon as BD comes out hahaha might be waiting a while.
    Reply
  • Noobz At Work - Wednesday, November 24, 2010 - link

    I already have the stock 7,200 RPM Hitachi HDD I pulled from my laptop which was replaced with odly enough, a drive that was reviewed here which would be the Seagate Momentus XT 320Gb. On the note of having the stock drive lying around, I am eventually going to get an enclosure that has an e-sata port so I can utilize the one I have on this laptop.

    Anyway, you still can't beat the write speeds of a good old fashioned mechincal drive, and not to mention another posters statment about how SSDs consume more power than a straight mechincal disk. That is why I never leave my battery in my computer, because I am sure the solid-state portion of the drive in my computer will eat away the battery more than usual.

    But, since I am partial to Kingston products(love my 16GB hardware-based 256AES locker+ data travler) Maybe next payday, I will order one just to have it for future use.
    Reply
  • LordOfTheBoired - Wednesday, November 24, 2010 - link

    Ummm.... no. The whole reason SSDs can be made to work without overhauling the host system architecture is that they DON'T draw power when off.
    They use flash RAM, just like thumb drives and SD cards and portable MP3 players and so on.
    As-is, drives usually have power physically removed when the system is turned off, so it's simply not POSSIBLE for them to draw power.

    As cool as it would BE if they used SRAM.... it's simply not going to happen. You'd have to redesign the power supply to provide constant power to the drives, and you'd lose your data every time the drive lost power.
    ...
    Unless it was battery-backed, but I suspect a hundred gigs of SRAM would require a significant battery.
    Reply
  • Chloiber - Thursday, November 25, 2010 - link

    "you still can't beat the write speeds of a good old fashioned mechincal drive"

    Err...yes you can.
    Reply
  • dqniel - Thursday, November 25, 2010 - link

    Uhh. I can't tell if you're being sarcastic or are just clueless. SSDs don't draw more power than mechanical drives even while under load, much less idle, they don't draw any power while the computer is powered off (they aren't using volatile memory), and the write speeds, read speeds, and most importantly access times are superior to mechanical drives. Reply
  • bahamakyle - Thursday, November 25, 2010 - link

    "That is why I never leave my battery in my computer, because I am sure the solid-state portion of the drive in my computer will eat away the battery more than usual."

    Nah man you're fine. Just save yourself the hassle and leave your battery in. Your SSD doesn't use any power when it's off. Unless of course you're joking haha.
    Reply
  • mpx - Sunday, November 28, 2010 - link

    Can you benchmark how well do these drives work for ReadyBoost? USB 2.0 flash drives and SSDs are quite slow (30MB/s is only 7,6k IOPS with 4kB pages), so ReadyBoost can't show it's potential. With this kind of drive, that has USB 3 interface it becomes much more interesting.

    How about other flash caching systems, like eBoostr or NVELO Dataplex?
    Reply

Log in

Don't have an account? Sign up now