Technology card
ShippingPerpendicular Magnetic Recording
PMR
Separate magnetic orientation, write assistance and track layout, then read the history and density claims in their original context.
Glossary entry on hesela.dev · JSON record · Magnetic anisotropy · Magnetic coercivity · Recording trilemma · Pole erasure · Exchange-coupled media
Related
On this page6 sections
In one minute
PMR records magnetization perpendicular to the disk surface; it specifies orientation, not whether tracks overlap or the writer uses heat.
Toshiba announced its MK4007GAL and MK8007GAH perpendicular-recording HDDs on 14 December 2004, with mass production planned for successive quarters of the fiscal year starting April 2005. Its retrospective milestones list commercialization in 2005. Announcement, planned production and a later historical summary are different evidence; none is the invention date of PMR.
A more specific introduction date appears in Toshiba's 5 June 2006 announcement: it places the 1.8-inch PMR HDD introduction in May 2005. This retrospective report strengthens the commercialization chronology, but does not supply an exact first-shipment day or confirm both models' production schedules individually. [1]
HGST's November 2007 white paper describes its own 2005 field test, 2006 product introduction and January 2007 terabyte-drive announcement. Its 2006 introduction must not be treated as the start of the entire industry's PMR history.
The magnetic recording trilemma connects readback signal quality, retention and writability. Keeping enough grains in a shrinking bit pushes grain volume down. At fixed anisotropy and temperature, this reduces the thermal reversal barrier; increasing anisotropy improves stability but makes writing harder. Evans et al. (2012) further distinguish overcoming the switching barrier from reliably selecting the intended state during a hot write.
Perpendicular anisotropy is also possible in HAMR media. Dong et al. (2016) fabricated FePt films on different intermediate layers and examined crystal texture, grain morphology and magnetic loops. A layer choice that improved grain isolation did not automatically preserve perpendicular anisotropy. Their mixed intermediate layer improved that balance: materials properties must be measured together, not inferred from a label.
A less visible design problem is pole erasure. Nakamoto et al. (2003) tested fabricated single-pole writers and linked erasure to remanent magnetization; reducing main-pole throat height suppressed it in their tested designs. That is evidence about a component mechanism, not an estimate of how often contemporary HDDs lose data.
Exchange-coupled composite (ECC) media add an internal grain-design choice. Magnetically soft and hard regions are coupled within a grain; this is not the exchange between adjacent grains. Kapoor, Shen and Victora's discretized single-grain model showed reversal starting in the soft region. Its comparison with a lower-anisotropy experimental sample also required different coupling: stronger is not a universal prescription. [9]
The reversal path matters as much as the material label. Suess et al. describe exchange-spring bilayers where nucleation in the soft region precedes domain-wall pinning at the interface. Their quasi-static model uses the larger of the nucleation and pinning fields for complete reversal. Graded media vary anisotropy through the thickness; predicted gains depend on the comparator, geometry and field protocol. Fabrication and head-field decay constrain ideal designs. [10]
Fast pulses and low damping can enable precessional reversal below the quasi-static switching field in the modeled bilayers. Suess et al. warn that changed angular sensitivity can compromise resistance to adjacent-track erasure. Static switching-field reduction alone is therefore insufficient to qualify a recording design. This is a model-dependent design tradeoff, not a measured drive error rate. [10]
Long simulated retention is not a drive-life measurement. Vogler et al. used forward flux sampling for same-geometry grains. The soft single-phase comparator matched the graded grain's coercive field, whereas the hard single-phase comparator had a similar energy barrier but a higher coercive field. These comparisons test different tradeoffs, not a universal retention advantage over single-phase media. They explicitly qualified the absolute escape rates as dependent on mesh size, even while defending the relative comparison. A retention headline without that caveat overstates what the simulation establishes. [11]
Compared to neighbors
Orientation, assistance and overlap are independent descriptions. Comparing PMR against HAMR as mutually exclusive categories hides the perpendicular anisotropy used by HAMR media.
| Property | Question answered | Not implied |
|---|---|---|
| PMR / longitudinal | Which way is magnetization oriented? | Whether tracks overlap |
| HAMR / unassisted writing | How is the medium made writable? | A specific track layout |
| CMR / SMR | Do recording tracks overlap? | A particular magnetic material |
| Exchange-coupled composite media | How are hard and soft regions coupled within a grain? | A universal coupling optimum or retention guarantee |
Uncertainty notes
- Reviewed 17 October 2026: corrected the scope of the 2006 milestone and dated the 2017 density estimate explicitly.
- Dong et al. studied deposited films, not commercial drive reliability; no replication or fleet result is claimed.
- The Nakamoto article's publisher abstract supports the component result. Its scanned full text was obtained but not text-reviewed; no threshold or numerical result is extracted.
- Reviewed 28 October 2026: added full-text ECC and graded-media research context; simulations and experimental comparisons retain their stated scope.
Sources
- 01Toshiba: 5 June 2006 announcement retrospectively dates 1.8-inch PMR HDD introduction to May 2005 · accessed 2026-10-23
- 02HGST: Perpendicular Magnetic Recording Technology, November 2007, pp. 1-4; hosted by Western Digital · accessed 2026-10-17
- 03Seagate: HAMR Technology, TP707.1-1712US, December 2017, pp. 1-3 · accessed 2026-10-17
- 04Toshiba: Perpendicular-recording HDD announcement, 14 December 2004 · accessed 2026-10-17
- 05Toshiba storage milestones: PMR commercialization, 2005 row · accessed 2026-10-17
- 06Evans et al.: Thermally induced error: density limit for magnetic data storage (2012), doi:10.1063/1.3691196 · accessed 2026-10-17
- 07Dong et al.: Columnar structured FePt films epitaxially grown on large lattice mismatched intermediate layer (2016), doi:10.1038/srep34637 · accessed 2026-10-17
- 08Nakamoto et al.: Single-Pole/TMR Heads for 100Gb/in2 Perpendicular Recording (2003), doi:10.3379/jmsjmag.27.124 · accessed 2026-10-17
- 09Kapoor, Shen and Victora: Effect of intragranular exchange on exchange-coupled composite media, J. Appl. Phys. 99, 08Q902 (2006), sections II-VI; doi:10.1063/1.2163850 · accessed 2026-10-28
- 10Suess, Lee, Fidler and Schrefl: Exchange coupled perpendicular media, arXiv:0805.1674v1 (2008), sections 1-3 and 4-5 · accessed 2026-10-28
- 11Vogler, Bruckner, Suess and Dellago: Calculating thermal stability and attempt frequency of advanced recording structures without free parameters, arXiv:1412.5057v1 (2014), pp. 3-5 · accessed 2026-10-28