Seagate floats 14 TB of storage in the helium-filled Exos X14

At this year's OCP Summit, Seagate teased a new drive for its storage-hungry enterprise clients. This hard drive adds a new tier of storage to its helium-filled Exos X line. Seagate says the 14-TB Exos X14 offers higher capacity than its siblings in the same form factor, while maintaining the same efficiency and performance.

 

Seagate's Exos X line of drives already includes 10-TB and 12-TB options, but the new 14-TB drive enables considerably more storage space per rack than the current models. The company isn't sharing hard numbers about the drive yet, but claims that the Exos X14 will boast the lowest power consumption of any comparable drive on the market while mantaining "leading" sustained transfer rates and random I/O responsiveness.

Seagate seems to have its eye on clients with strict security requirements. The Exos X14 offers hardware encryption and should meet the requirements for the U.S. government's FIPS 140-2 standard and the international ISO/EIC 15408 spec.

Currently, Seagate is sampling the drive to select customers including Chinese search provider Baidu. The company says that the Exos X14 should be broadly available starting this summer. Seagate hasn't listed a price yet for the capacious drive, but it's unlikely that its primary customers will be buying one at a time.

Comments closed
    • freebird
    • 2 years ago

    I know something else that “floats” and begins with an ‘S’ also… ;D

    Seagate might be fine, but I have had better luck and performance with HGST & Toshiba the past few years.

    Now if they had an ‘X-15’ Model like this,
    [url<]https://www.youtube.com/watch?v=_H2he0UUXW4[/url<] I might be interested.

    • BIF
    • 2 years ago

    Okay, dumb question:

    I thought there was a severely limited supply of helium in the world, and it’s shrinking, not growing. Why don’t these cost a million dollars each?

      • JustAnEngineer
      • 2 years ago

      Helium is almost entirely tied up in the same formations that hold natural gas deposits. With the wild success of hydraulic fracturing, there’s been so much natural gas extracted from the earth in recent years that the helium repositories are overflowing.
      [url<]https://en.wikipedia.org/wiki/National_Helium_Reserve[/url<]

        • BIF
        • 2 years ago

        Oh wow, I didn’t know that. Thank you.

      • mikewinddale
      • 2 years ago

      Also, the US government – which held the Helium Reserve – began selling off the helium at below-market prices. This encouraged over-consumption.

      This may explain why we see things like helium-filled party balloons. Every party balloon will eventually release its helium into the atmosphere, and it will be difficult if not impossible to regain that lost helium to use in an MRI machine.

      So basically, governments gives away a resource below-cost, and the resource gets wasted. Same thing happens around the world with water: governments often subsidize water for farmers, and then we wonder why growing avocados in California is depleting the aquifers.

    • ronch
    • 2 years ago

    And here I am still running a 1TB Hitachi from 2011 with around 30GB of free space. I’ll probably need to look around for a new, bigger drive soon.

    • just brew it!
    • 2 years ago

    [quote=”Seagate”<]Seagate creates space for the human experience by innovating how data is stored, shared and used.[/quote<] "Creates space for the human experience"? WTF...

    • Welch
    • 2 years ago

    Well, with how data sizes keep ballooning, I see no way but up in size. Glad to see this series of drives take off. Maybe it can help Seagate keep their business afloat.

    • Sargent Duck
    • 2 years ago

    2TB SSD’s are starting to permeate the consumer space and I imagine it won’t be long before we see 4 and 8TB SSD’s.

    So to Seagate, I say *yawn*.

      • designerfx
      • 2 years ago

      100% this. I was debating building a NAS but for longevity and a host of reasons I may as well wait until SSD’s expand into higher storage capacities.

        • Blytz
        • 2 years ago

        Just get good value for TB drives now and upgrade later, otherwise you’ll always be waiting for the next storage capacity

      • just brew it!
      • 2 years ago

      I guess you missed the word “enterprise” in the article. You are not the target market for this drive!

      When you’re dealing with Cloud-scale bulk data storage applications (we’re talking storage farms where a petabyte is considered “small”), cost per byte tops the list of priorities. SSDs still can’t compete with spinning rust on a cost basis (and despite predictions to the contrary, probably won’t any time soon).

        • BIF
        • 2 years ago

        +3 to this.

        But there’s another great reason for having such a huge capacity drive, and that’s for backups.

        I’d love to have a few 14 TB spinners available in a NAS-based enclosure for a dedicated personal disaster recovery plan. That way, I’d be able to back up numerous cycles of all of my computer hard drives onto one huge spinner, and then only toss that one drive into my offsite vault (my locked desk drawer at work).

        Unfortunately for me, NAS device and drives will probably be cost-prohibitive. I’d need at least 3 of them to put into a rotation. One current, one prior, and one always offsite.

      • joekraska
      • 2 years ago

      Large archival operations are dominated by $/GB, and even accounting for the extra space used, large spinning media still has a place. For the time being.

    • JustAnEngineer
    • 2 years ago

    Is this giant drive infected with shingles?

      • Krogoth
      • 2 years ago

      I don’t think so. It is just high on Helium.

      • DancinJack
      • 2 years ago

      PMR

      • Waco
      • 2 years ago

      No 3.5″ drives going forward (10 TB+) will be drive-managed SMR. All will be explicitly host-managed.

    • sweatshopking
    • 2 years ago

    that’s a lot of data to lose by going Seagate…

      • Waco
      • 2 years ago

      Yeah, those AFRs competitive with the industry just aren’t good enough. Sigh.

      • Kaleid
      • 2 years ago

      Aren’t most of the bad drives the 4TB versions?

        • just brew it!
        • 2 years ago

        [url=https://en.wikipedia.org/wiki/ST3000DM001#Class_action<]3TB, actually[/url<]

      • albundy
      • 2 years ago

      well, its either them or WD/HGST. thats it. there is no competition really.

      • just brew it!
      • 2 years ago

      Enterprise-scale storage (the target market for drives like this) always employs redundancy. Simple mirroring or even RAID-5/6 is costly and/or insufficiently robust at petabyte (and beyond) scale though. These days large-scale storage systems tend to use some form of [url=https://en.wikipedia.org/wiki/Erasure_code<]k/n erasure coding[/url<], which can be applied across individual drives, entire servers/clusters, or even geographically dispersed datacenters.

        • joekraska
        • 2 years ago

        In fact the largest scale operations have all moved to triple redundancy (e.g., S3), and to be damned with the costs.

          • just brew it!
          • 2 years ago

          Use of erasure coding across datacenters can allow you to get some of the benefit of triple redundancy, without actually needing to duplicate all of the data 3 times. Some large operations are in fact doing this (not sure about S3). You can lose an entire datacenter and still recover your data. Of course it means if you lose 2 datacenters out of 3 you’re still screwed; but if they’re sufficiently dispersed geographically that risk/cost tradeoff may be considered acceptable.

    • DPete27
    • 2 years ago

    Hmm, with SSDs achieving 100TB these days, 14TB is feeling kinda cramped….

      • Wirko
      • 2 years ago

      And you haven’t seen helium-filled SSDs yet!

        • the
        • 2 years ago

        Of course not, they have already floated away.

      • Blytz
      • 2 years ago

      Hear to hoping it hammers the spinning rust division into action

      • Waco
      • 2 years ago

      Capacity total matters a lot less when you’re talking 3 cents per GB compared to 40+ cents per GB…

      It’s attractive, sometimes, if you are completely density or power constrained…but at a 10x+ cost deficit it’s easy to justify the spinning rust.

        • JustAnEngineer
        • 2 years ago

        Storage is purchased by the TB in this decade. GB are passé. Consumer hard drives are in the neighborhood of $25/TB, which is still almost an order of magnitude less than the $230/TB of SSDs.

          • Waco
          • 2 years ago

          I buy petabytes per purchase and still use cents per Gold habits are hard to break.

            • just brew it!
            • 2 years ago

            I just realized that a single 36-drive 4U server chassis can hold over half a petabyte using these drives…

            • Waco
            • 2 years ago

            If you can tolerate deeper chassis (and lots of lb/sq foot), the 96 drive JBOD/server combos can put nearly 1.4 PB raw into 4U.

            I hate embedded servers, though, so I’ll always sacrifice a bit of density to use whitebox servers to save a few bucks. You can’t realistically put 10-12 of these in a 40/48U rack in any normal datacenter due to weight restrictions anyway.

            Our (LANL MarFS build) standing design for a “rack unit” is 6 5U 84 drive JBODs with 3 2U servers. It’s under typical datacenter weight limits and allows for a lot of flexibility on server choice for price versus performance trade-offs. We could jump to denser JBODs at the next deployment, but embedded servers just don’t work at scale if you care about cost and have a normal floor.

            • just brew it!
            • 2 years ago

            Yeah, some of our customers have 96-drive units. I haven’t personally laid eyes/hands on one though; the densest ones I’ve routinely dealt with hold 48 drives. And yes, they’re f’ing heavy; maneuvering a fully loaded one in/out of a rack is challenging!

            There are also some stories (from before I joined the company) about people damaging racks and floors before everyone was aware of the weight issues you start getting into when you stuff racks full of these things.

            • Waco
            • 2 years ago

            Oh yeah, you don’t want to move one loaded. A lot of the denser JBODs specify that movement with drives loaded violates any warranty…

            • just brew it!
            • 2 years ago

            I’ve moved and racked the 48-drive ones fully loaded. You want to have something like [url=https://www.northerntool.com/shop/tools/product_200612504_200612504<]this[/url<], and (ideally) an assistant. Based on that experience, I can see how doing that with a 96-drive enclosure would be highly inadvisable, if not impossible!

            • Waco
            • 2 years ago

            We use electric lifts with 300+ pound capacity. I still won’t move (or ask anyone to move) a fully loaded JBOD with 60+ drives.

            • just brew it!
            • 2 years ago

            It also doesn’t help that the flagship capacity drives have a ridiculous number of platters in them. The weight difference between the high capacity models and typical consumer drives is pretty significant.

Pin It on Pinterest

Share This