Seagate on Monday disclosed that it had begun testing the industry’s first HAMR hard drive intended for evetualy commercial release. With a capacity of 16 TB, the HDD is being used primarily for internal tests to prepare for its high-volume launch and deployment in actual datacenters in the future. Separately, Seagate announced plans to introduce HAMR-based hard drives with a 20 TB capacity in 2020.

Seagate’s 16 TB Exos HDD featuring heat-assisted magnetic recording technology are drop-in compatible with existing servers and datacenters, which essentially means that their power consumption is 12 W or below. The hard drive is helium filled, but Seagate does not disclose the number of platters the HDD uses.

Right now, Seagate is testing its 16 TB Exos HDDs in a variety of benchmarks internally. Next the company will provide the drives to its customers, and once those customers qualify them Seagate will finally and officially launch the product.

“These are the same tests that customers use to qualify every new drive, including power efficiency tests, sg3_utils utilities that test SCSI commands to devices, standard smartmontools utility programs that will enable customers to characterize and compare HAMR drives in their environment right next to PMR drives, and several four-corners tests of reads, writes, random, sequential and mixed workloads,” said Jason Feist, Seagate’s senior director of enterprise product line management.

Related Reading:

Source: Seagate

Comments Locked


View All Comments

  • MrSpadge - Wednesday, December 5, 2018 - link

    Just summing up additional 200 mW for a huge number of drives is meaningless. The interesting comparison would be against comparable 14 TB HDDs without HAMR. There the power consumption would be definitely be higher as you simply need more drives. Helium filled drives may challange the first generation of HAMR drives in TCO, though. Long term both technologies cold be merged.

    Side note: W is already "energy per time", so W/s actually makes it wrong.
  • ikjadoon - Wednesday, December 5, 2018 - link

    Whatever happen to WD's MAMR? WD deadlines were all kinds of wrong. WD thought QLC wouldn't arrive until 2022 // HAMR had "unknown manufacturability" // MAMR was ready for "Production in 2019" // charts showing MAMR with a significant lead in deployment over HAMR.

    Anandtech repeated WD's claims over a year ago:

    Crickets from WD since then.
  • jjj - Wednesday, December 5, 2018 - link

    WDC is sampling 16TB with 8 platter and they expect 20TB before 2020.
    slide 32 (large PDF)

Log in

Don't have an account? Sign up now