{
  "slug": "track-density",
  "term": "Track density",
  "aliases": [
    "TPI",
    "tracks per inch"
  ],
  "definition": "Track density is the number of recording tracks per unit distance measured across the tracks.",
  "note": "Tracks per inch (TPI) describes the cross-track dimension. For uniformly spaced track centers, TPI is the reciprocal of track pitch measured in inches. It is distinct from bits per inch along a track. [tape-density]\n\nNarrower pitch places more tracks in a given width. SMR uses overlapping writes to leave readable tracks narrower than the write head's footprint, so write-head width alone does not determine the final track pitch. [smr-models]\n\nHigher TPI at fixed BPI increases local areal density. At fixed disk radius and RPM it does not, by itself, increase the idealized bits passing a single head per revolution. Peak and average specifications must retain their original scope. [hdd-card] [density-manual]\n\nReader width is not track spacing. Furrer et al.'s SrFe study combines a 29 nm reader with a calculated 56.2 nm track width, giving about 452 ktpi. Section V adds a positioning margin using 2 x sqrt(2) x 3 x sigma_PES + reader width, with sigma_PES rounded to 3.2 nm. Using 29 nm alone would imply about 1.94 times the track density. This is the authors' prototype design calculation, not a universal pitch rule. [tape-servo]\n\nThe recording test used an 8.7 micrometre writer, while servo tracking was evaluated separately. The 56.2 nm figure combines these results through a spacing model; it is not the written width of those experimental data tracks. [tape-servo]",
  "status": "shipping",
  "advantages": [],
  "limitations": [],
  "related": [
    "areal-density",
    "linear-recording-density",
    "hdd",
    "smr"
  ],
  "supersedes": [],
  "supersededBy": [],
  "firstSeen": null,
  "lastSeen": null,
  "citations": [
    {
      "id": "tape-density",
      "title": "Furrer et al.: 201 Gb/in2 Recording Areal Density on Sputtered Magnetic Tape, IEEE Transactions on Magnetics (2018)",
      "url": "https://research.ibm.com/publications/201-gbinlesssupgreater2lesssupgreater-recording-areal-density-on-sputtered-magnetic-tape",
      "accessed": "2026-10-19"
    },
    {
      "id": "smr-models",
      "title": "Zoned Storage: Shingled Magnetic Recording Hard Disks, overview and interface implementations",
      "url": "https://zonedstorage.io/docs/introduction/smr",
      "accessed": "2026-10-19"
    },
    {
      "id": "hdd-card",
      "title": "Hesela: Hard Disk Drive, density and transfer-rate interpretation",
      "url": "https://hesela.com/tech/hdd/",
      "accessed": "2026-10-19"
    },
    {
      "id": "density-manual",
      "title": "Seagate: Enterprise Capacity 3.5 HDD v4 SAS Product Manual, 100726011 Rev. A, May 2014, sections 3.5 and 4.1",
      "url": "https://www.seagate.com/www-content/product-content/enterprise-hdd-fam/enterprise-capacity-3-5-hdd/constellation-es-4/en-us/docs/100726011a.pdf",
      "accessed": "2026-10-19"
    },
    {
      "id": "tape-servo",
      "title": "Furrer et al.: 317 Gb/in2 Recording Areal Density on Strontium Ferrite Tape, author preprint v1, sections III, IV and V",
      "url": "https://doi.org/10.36227/techrxiv.13379594.v1",
      "accessed": "2026-10-19"
    }
  ],
  "updated": "2026-10-19",
  "provenance": [
    "Furrer et al. (2018), IBM Research abstract, links 103 nm track width with 246.2 ktpi in a prototype tape experiment; no product-capacity extrapolation is made. [tape-density]",
    "The author preprint was read beyond its abstract. Its 317.3 Gb/in2 result combines measured linear density with calculated cross-track spacing; section III.2 extrapolates the very low post-decoding error rate under an uncorrelated-byte-error assumption. [tape-servo]"
  ]
}