At the core of customer-centric innovation is a simple truth: customers don’t want technologies, they want outcomes. In data storage, those outcomes increasingly come down to one thing above all others: scalable capacity delivered with predictable economics and reliability.

Areal density drives HDD economics

One of the most critical outcomes customers need from storage is total cost of ownership (TCO)—not just at deployment, but across years of operation at massive scale. There are many factors involved in TCO, one of the most important being drive capacity, which impacts storage density (TB/in³) and power efficiency (W/TB).

Drive capacity can be achieved in many ways, such as mechanical designs that increase the number of disks per 3.5” HDD. But if we don’t want to or can’t add disk area via mechanical means, we can look to increasing areal density capability (ADC), a measure of how much data can be stored per square inch on a magnetic disk.

The history and future of ePMR

Over the past decade, WD has systematically expanded the limits of perpendicular magnetic recording (PMR) technology, delivering some of the most consistent and predictable capacity gains in the industry. This includes being the first to bring energy-assisted magnetic recording (EAMR) to market with ePMR technology, which has expanded drive capacity today to 30TB in CMR and 40TB in UltraSMR, the largest ePMR drives in the industry. This gives customers a consistent and predictable capacity upgrade path—one that offers seamless qualification and a rapid deployment ramp.

WD continues to extract every viable gain from ePMR, with a roadmap to capacities of 60TB. However, as the fundamental limits of physics and magnetics are approached, the pace of areal density improvement is inevitably slowing. New technologies are required to achieve our customers’ capacity demands of the future. To move beyond these limits and deliver the next step-function increase in capacity, the industry must adopt a fundamentally different recording approach: which brings us to heat-assisted magnetic recording (HAMR), which will enable us to extend capacities up to or above 100TB.

The superparamagnetic limit and the magnetic recording trilemma

The push to higher areal density involves balancing multiple competing forces to overcome the superparamagnetic limit, a fundamental physical limit that prevents bits from being made smaller without risking data integrity. This is known as the magnetic recording trilemma:

• To increase areal density, the size of the written bits must shrink. As this happens, the amount of energy needed to flip the magnetization drops.
• To keep the magnetization of bits from flipping due to thermal energy, engineers must use media materials with higher anisotropy (magnetic resistance to flipping) to avoid unwanted changes.
• As the writer defines the size of the written bit, smaller written bits require smaller writers, and smaller writers generally generate less magnetic field. However, to write data on high-anisotropy media, the magnetic field generated by the write head must increase.

With conventional recording, the trilemma is becoming more difficult to solve. The smaller geometry of the writer is difficult to optimize further, and it already uses materials capable of generating the strongest magnetic field and small writer size. Thus, further increasing the media anisotropy becomes a more difficult option with conventional recording.

While ePMR has a future, new technologies are needed to achieve the huge capacities the market increasingly demands.

HAMR opens the door to higher anisotropy

One way to address the trilemma is to apply additional energy to temporarily modify attributes of the media, making it easier to write with conventional field strength. HAMR allows for higher-anisotropy media, and a corresponding reduction in minimum bit size and increase in areal density.

HAMR accomplishes this via heat. When the media temperature increases at the spot to be written, the anisotropy decreases, and that region of media is temporarily easier to write. When it cools down, the anisotropy increases, and the media becomes harder to write and more thermally stable.

Thus, by using a laser to heat only the area needing to be written, the drive can use a conventional magnetic write head and still effectively write higher-anisotropy media than conventional PMR can write. When that area cools, each grain of the media stabilizes in the desired magnetic polarity.

With the addition of HAMR, the Advanced Storage Research Council (ASRC) projects¹ that areal density growth can return to a ~20% CAGR—restoring the trajectory required to support AI-scale data growth.

WD’s customer-centric HAMR strategy

While this all sounds amazing—and it is!—HAMR is a complex and challenging technology. To achieve its promise, its development and deployment must be customer-centric. Critically, cost and reliability are two customer needs that HAMR technology stresses. HAMR adds bill of material (BOM) cost through two primary components: glass substrate media (impacting all 11 disks) and the laser and optical transducer (impacting all 22 heads).

When it comes to reliability, the ability to make the lasers in the drives’ write heads last and achieve the MTBF and workload ratings that our customers expect is non-negotiable. WD believes the optimal inflection point for HAMR—where capacity gains outweigh added cost and reliability risk—emerges at 40TB and above.

For this reason, WD has a robust roadmap for continued innovation in ePMR capacity gains. The technologies we have used to bridge to HAMR, such as the triple stage actuator (TSA), OptiNAND®, UltraSMR, and our 11-disk mechanical platform, will allow us to maximize capacity gains when we carry them forward to HAMR. The introduction of HAMR does not foretell the end of ePMR. Instead, it expands the range of options customers have to align technology choices with their workloads, economics, and deployment timelines.

ePMR and HAMR: A parallel innovation roadmap

HAMR is a necessary next step for the HDD industry to achieve the massive capacities needed to continue as the indispensable infrastructure of the AI-driven data economy. WD has been innovating in HAMR while continuing to supply the technologies and capacities needed to not release HAMR before it was truly ready. As customers are now qualifying WD HAMR HDDs, we are working to ensure that they have a smooth, predictable transition to this revolutionary technology.

By advancing ePMR and HAMR in parallel, WD helps ensure that our customer roadmaps—not vendor constraints—define the pace of innovation. That discipline is what allows us to deliver the right technology, at the right time, for the right customer need.

*https://ieeexplore.ieee.org/document/9918580