Western Digital has introduced its new top-of-the-range air-filled hard drive for enterprises that use storage servers, RAID arrays, and distributed storage. The Ultrastar DC HC330 HDD expands the capacities that Western Digital offers for this market segment to 10 TB and also introduces the company’s Media Cache Plus technology that promises to improve write performance by up to 40% in certain cases.

The Western Digital Ultrastar DC HC330 is the latest addition to the company’s Ultrastar DC HC300-series family that includes 4, 6, and 8 TB hard drives (that were originally introduced under different model numbers). While these hard drives are classified as ‘datacenter’ devices, they are aimed at enterprises that need loads of storage space and high performance, but not those operating large datacenters that require maximum capacities. All the HDDs in the lineup are based on the company’s enterprise-class platform that uses conventional magnetic recording (CMR, which refers to perpendicular magnetic recording in this case) featuring appropriate reliability and durability enhancements (e.g., bottom and top attached motors, RAFF technology to protect against vibration, dual-stage micro actuator to improve positioning, etc.).

The Ultrastar DC HC330 10 TB air-filled hard drives rely on six 1.66-TB CMR platters (the highest-capacity commercially available platters in the industry today), down from seven 1.42 TB PMR disks in case of the helium-filled Ultrastar He10. The HDDs use a motor featuring a 7200 RPM spindle speed, a 256 MB buffer, and will be offered with a SATA 6 Gbps or a SAS 12 Gbps interface as well as 4 KB and 512E sectors.

Usage of high-capacity platters enabled the Ultrastar DC HC330 10 TB to increase its sustained transfer rate up to 273 MB/s (faster than cheap SSDs when it comes to large data transfers). In addition, the new hard drive features non-volatile Media Cache Plus technology that boosts random write performance by up to 40% when the queue depth is low or the transfer length is short.

In theory, reduction of the number of platters should have reduced the power consumption of the Ultrastar DC H330 compared to the Ultrastar He10, yet it is not the case as the drive is air-filled, not helium-filled. The new drive consumes up to 9.2 W of power in operating mode, up from 6.8 W in case of its predecessor. This does not mean that the new HDD is necessarily worse than its 10 TB predecessor. Firstly, it is cheaper to manufacture, and secondly, since it is faster in general use-cases as well as in certain write-intensive scenarios, it may still be power efficient enough for most use-cases (especially keeping positioning of the product). On the other hand, even when in idle, it consumes 3 W more when compared to its predecessor (8 W vs. 5 W), which might be a problem for companies like Amazon or Netflix that use millions of HDDs across tens of large datacenters, but shouldn't represent too much of a big issue for companies employing smaller deployments.

HGST Ultrastar Ultrastar DC HC300-Series General Specifications
  Ultrastar DC HC 310 Ultrastar DC HC 320 Ultrastar  DC HC330
Capacity 4 TB, 6 TB 8 TB 10 TB
RPM 7200 RPM
Interface SATA 6Gbps SAS 12 Gbps SATA 6 Gbps SAS 12 Gbps SATA 6 Gbps SAS 12 Gbps
Format: Sector Sizes 4Kn 4096 4096, 4104
4160, 4224
4096 4096, 4104,
4160, 4224
4096 4096, 4104,
4160, 4224
512n 512 - -   -
512e 512 512, 520, 528 512 512, 520, 528 512 512, 520, 528
DRAM Cache 256 MB
Helium-Filling No
Sustained Transfer Rate Up to 255 MB/s (4Kn)
Up to 233 MB/s (512n)
Up to 273 MB/s
Average Latency 4.16 ms
Seek Time (read/write) 8/8.6 ms
Acoustics 2.9/3.6 Bels 3.4/3.8 Bels
Power Rating Idle 5.9 W 7 W 7 W 8 W 8 W
Operating 7 W 12.1 W 8.3 W 13.4 W 9.2 W
MTBF 2 million hours
Warranty 5 Years

Covered by a five-year warranty, Western Digital’s Ultrastar DC HC330 10 TB HDDs will be available to OEM, cloud, and enterprise customers in September. Pricing will depend on volumes, amongst other factors.

Related Reading:

Source: Western Digital

Comments Locked

8 Comments

View All Comments

  • dgingeri - Wednesday, July 24, 2019 - link

    So, mechanical hard drives have finally gone beyond the transfer rate of 3Gb/s SATA.
  • Samus - Wednesday, July 24, 2019 - link

    Mechanical hard drives have saturated the transfer rate of SATA3 since they moved to 256MB buffers years ago, they just don't do it for more than a second or two. This drive will actually be close to SATA3 limits during short burst transfers - the same way SSD's have been before going to NVMe, due to its large buffer AND incredibly high sustained transfer rate. This is yet another reason you see SAS used on enterprise drives, especially when a RAID is used.

    But I get the joke you're implying, it's just important to keep in perspective the majority of workloads for these drives is burst, not sustained transfers ;)
  • MrSpadge - Wednesday, July 24, 2019 - link

    Yes, that cache works for shorts bursts. But the chance of hitting the cache for reads is very low, especially if you consider that the OS already uses GBs of "free" DRAM as read cache.
  • deil - Wednesday, July 24, 2019 - link

    I still wonder why we don't have mixed drives.
    10 TB HDD for everything 1MB or bigger.
    512 GB SSD for everything 1MB or smaller and indexing HDD.
    Small does not clog IOPS on HDD, big does not fill SSD.
    search time is at SSD speed, capacity is enough for most usecases.
  • buxe2quec - Wednesday, July 24, 2019 - link

    Something similar is called ZFS and caches on a SSD (L2ARC) the files you actually use, so that you can get 500 MB/s (or whatever your SSD does) even on big ones, without wasting SSD for the many but rarely used small ones.

    Not exactly what you asked, but AFAIK the best compromise so far.
  • igavus - Wednesday, July 24, 2019 - link

    Also add compression to that and the effective read speeds are really magical with NVME ;-)
  • CrazyElf - Wednesday, July 24, 2019 - link

    Technically we have solid state hybrid drives (SSHD).
  • tommo1982 - Sunday, July 28, 2019 - link

    I wonder. How about moving whole MBR to MMR. It's nonvolatile, fast and you could save a lot on latency when heads don't have to seek where the data is stored.

Log in

Don't have an account? Sign up now