A tape drive costs more up front than buying disk for a small archive - the math only turns tape's way once you are storing enough that its cheaper media cost has paid back the drive. This page computes that crossover size live from our own catalogue: the drive's entry cost divided by the gap between disk $/TB and tape media $/TB. Below that size, disk is genuinely cheaper on these inputs, and this page says so rather than selling tape regardless.
Computed from live catalogue prices - nothing on this page is a fixed number.
break-even TB = drive entry cost ÷ (disk $/TB − tape media $/TB, native)
Every term on the right is pulled live from our own catalogue at render time - the drive's entry cost (drive plus a compatible host bus adapter plus a cleaning cartridge, the same basket the drives hub prices), the cheapest in-stock native $/TB for that tape generation's media, and the current HDD floor. None of the three is hardcoded on this page; if disk stops being more expensive per TB than tape media, the denominator goes to zero or negative and this page says the archive never pays the drive back, rather than dividing by a number that shouldn't be divided by.
We only compute this where this site tracks BOTH a live drive price and a live media price for the same generation - currently LTO-8 and LTO-9. LTO-5 and LTO-4 drives are in our catalogue, but we do not stock LTO-5 or LTO-4 media, so there is no live tape $/TB to divide against for those generations and we do not estimate one.
| Generation | Drive entry cost | Tape media, $/TB native | Disk reference | Disk $/TB | Break-even |
|---|---|---|---|---|---|
| LTO-8 | $4966 | $6.42 | New-drive floor | $15.00 | 578.5 TB (48.2 cartridges) |
| LTO-8 | $4966 | $6.42 | Refurb-drive floor | $14.99 | 579.2 TB (48.3 cartridges) |
| LTO-9 | $10493 * | $4.89 | New-drive floor | $15.00 | 1037.8 TB (57.7 cartridges) |
| LTO-9 | $10493 * | $4.89 | Refurb-drive floor | $14.99 | 1038.8 TB (57.7 cartridges) |
* The LTO-9 drive we track is external and needs a host bus adapter with an external SAS-3 port - our HBA catalogue is internal-port only, so no compatible card is priced into this figure. Its true entry cost, and therefore its true break-even size, is higher than shown.
On today's prices, an LTO-8 setup needs roughly 578.5TB of archive before its media savings have paid back the drive against new-disk $/TB - below that, buying disk is cheaper. If your archive is smaller than the relevant row above, this page is telling you plainly that tape does not pay yet at your scale - the honest answer routes you back to disk and refurbished disk, not toward a tape purchase this math cannot justify.
Every LTO generation is marketed on a compressed capacity that assumes a 2.5:1 ratio the LTO Program states in a footnote - already-compressed data such as video or photos will not compress again. This page uses native $/TB on the tape side for the same reason the media hub does: a buyer sizing an archive on the marketed number will buy too few cartridges, which would make tape look like it pays back sooner than it actually does for real-world already-compressed data.
Tape drives & the live entry-cost basket · LTO tape media $/TB · Datacenter storage TCO report · New disk · Refurbished disk.
The archive size, in terabytes, at which the money you'd have spent buying disk instead equals the money you spent on a tape drive plus enough tape media to hold the same data. Below that size, tape's cheaper media hasn't yet earned back the drive's up-front cost, so disk is cheaper overall. Above it, tape is cheaper overall on media cost alone. It is not a recommendation - it is one number, computed live, with everything it leaves out stated below rather than hidden.
Because they give genuinely different answers, and averaging them would hide that. Refurbished enterprise drives are typically the cheapest $/TB this site tracks, which raises tape's break-even point - it takes a bigger archive for tape to catch up against the cheapest disk available. New-drive $/TB is higher, so tape catches up sooner against it. Both are real, live numbers; neither is the "right" one to use - it depends on whether refurb disk is actually an option you'd consider.
No, and this page says so directly rather than implying it. This is a capacity break-even on two numbers: drive entry cost and media $/TB. It excludes power, labour, restore time, drive redundancy, and every operational cost a real deployment carries on both sides. See the full report linked below for a fuller treatment, and see the exclusions section on this page before treating this number as a business case.
Because we do not sell LTO-5 or LTO-4 media, so we have no live tape $/TB figure for those generations to divide against - and we will not invent one. The table only covers generations where this site tracks both a hardware price and a media price: currently LTO-8 and LTO-9. See the drives hub for the LTO-5 drives we do track.
Then your break-even math is entirely different, and lower: the entry cost this page charges against tape is the drive itself, which you have already paid for. Re-run the same formula with your sunk drive cost at $0 and it becomes a comparison of media $/TB alone, which tape already wins on at every generation we track - see the media hub's own $/TB table.
Yes - the entry cost in this page's math is the same live basket priced on the drives hub: drive, a compatible host bus adapter (where our catalogue has one), and a cleaning cartridge rated for that generation. It does not add a second full cartridge cost beyond the one included in the drive-hub basket - see the exclusions section.
Specification claims on this page were checked against the pages below on 4 August 2026. Where a widely-repeated figure did not survive that check, we report what the source says now and say so in the text rather than repeating the familiar number.