Now, newly-unearthed patents appear to show that the company may want to (re) add tape to complete its current media portfolio after the sun-setting of its Arkeia product range a few years ago.
The company was recently assigned a number of patents mentioning “tape embedded drive” in recent years:
- 11393498 (PDF) (head assembly with suspension system for a tape embedded drive)
- 20200258544 (PDF) (Tape embedded drive)
- 11081132 (PDF) Tape embedded drive with HDD components
- And a few more
What this alludes to is the intriguing possibility of getting the basic components of a tape drive merged with the actual tape media in a bid to reduce the inherent environmental and technological complexity of tape libraries as well as improving the access time by at least one order of magnitude.
Having the read and write heads closer to the media in an enclosed form factor is nothing new. That’s what hard disk drives do and what others, most notably Iomega with its Zip Drive, has tried to do in the past. Western Digital’s patent suggests adopting a standardized form factor, 2.5-inch or 3.5-inch, for easier adoption by data centers and hyperscalers.
The cost factor
An embedded tape would still be more expensive than a normal one (LTO-9 tapes retail for about $130 a pop) because of the extra electronics but you don’t need a tape drive to get started. As long as it sits somewhere between tapes ($4 per TB) and enterprise hard drives ($20 per TB), there will be a significant market for it.
A standard LTO (Linear Tape-Open) tape is 102 x 105 x 21mm while your average 3.5-inch hard drive is about 147 x 101 x 26mm while weighing about a quarter of the weight. LTO-9 has a compressed capacity of 45TB (18TB uncompressed) with the next generation – likely to come in the second half of this decade – doubling in capacity (obviously, there might be some adjustments as it was the case going from Gen 8 to Gen 9).
A sealed LTO-based tape drive would likely be lighter, cheaper, consume/dissipate less power but also have more onboard compute capabilities than your standard hard disk. Thicker and wider tape reels would also allow far bigger capacities (LTO-9 uses 1Km of tape with a 5.2µm thickness and 12.65mm width).
Western Digital is uniquely positioned to convert this into a workable reality especially as it can use its expertise in hard drive components. That new tape could, all things considered, use a similar PCB and interface as an enterprise hard disk drive; it doesn’t need to have the traditional tape look.
Perhaps the biggest challenge is to take this novel approach to tape to the LTO consortium, an organization that oversees the development of LTO and comprises IBM, HPE and Quantum, all of whom may have different commercial strategies that require having an expensive drive and cheap tapes.