| Pebioctets (Pio) | Yottaoctets (Yo) |
|---|---|
| 1 Pebioctet | 0.00000000112589990684 Yo |
| 2 Pebioctets | 0.00000000225179981369 Yo |
| 3 Pebioctets | 0.00000000337769972053 Yo |
| 4 Pebioctets | 0.00000000450359962737 Yo |
| 5 Pebioctets | 0.00000000562949953421 Yo |
| 10 Pebioctets | 0.0000000112589990684 Yo |
| 20 Pebioctets | 0.0000000225179981369 Yo |
| 25 Pebioctets | 0.0000000281474976711 Yo |
| 50 Pebioctets | 0.0000000562949953421 Yo |
| 100 Pebioctets | 0.000000112589990684 Yo |
| Reference | Pebioctets (Pio) | Yottaoctets (Yo) |
|---|---|---|
| A plain text message (160 characters) | 1.42109 × 10-13 Pio | 1.6 × 10-22 Yo |
| A three-minute MP3 | 0.00000000266454 Pio | 3 × 10-18 Yo |
| A smartphone photo | 0.00000000355271 Pio | 4 × 10-18 Yo |
| A high-definition film | 0.00000355271 Pio | 4 × 10-15 Yo |
| A dual-layer Blu-ray disc | 0.0000444089 Pio | 5 × 10-14 Yo |
The pebioctet is a unit of digital information equal to 1,125,899,906,842,624 octets, which is 1,024 tebioctets. Its symbol is Pio. It is the binary counterpart of the petaoctet, and the two now differ by 12.6 per cent.
This is the unit in which filesystem limits are written. Several widely deployed filesystems have maximum file or volume sizes that are exact powers of two, and those limits fall in this range: sixteen tebioctets for one, a pebioctet for another, eight pebioctets for a third. The numbers are not choices made by anyone; they follow from how many bits the filesystem uses to number its blocks.
Large scientific and commercial storage systems are also sized here. A research computing centre may run a filesystem of several pebioctets across thousands of drives, and the tools that manage it report in pebioctets because the underlying allocation is binary throughout. Converting those figures to decimal units for a report introduces an error of an eighth for no benefit.
A pebioctet is roughly a thousand large consumer hard drives, or the complete video output of a broadcaster over many years. It is also close to the size of the largest single scientific archives, though those are typically stated in petaoctets by press offices and in pebioctets by the engineers who run them, which produces two different numbers for the same building.
Distinguishing the two is not pedantry at this level. A twelve per cent difference on a purchase of many pebioctets is a large sum of money, and a capacity plan built on the wrong unit will be wrong by more than the margin most plans allow. This is precisely the situation the IEC prefixes were introduced to prevent.
The symbol Pio is used in filesystem documentation, in storage system manuals and in standards. Where a source writes PB without saying which it means, the difference is an eighth, and the honest thing for a reader to do is treat the figure as approximate rather than assume one convention or the other.
One pebioctet equals 1,024 tebioctets, 1,125,899,906,842,624 octets, 8 pebibits, or about 1.126 petaoctets.
The yottaoctet is a unit of digital information equal to a thousand zettaoctets, or a septillion octets — a one followed by twenty-four zeros. Its symbol is Yo. Nothing that exists is measured in yottaoctets; the unit describes a quantity of data the world has not yet produced.
The total data held by humanity is currently estimated in the low hundreds of zettaoctets, so the world is a small fraction of the way to its first yottaoctet. On the growth rates of the last two decades, that threshold would be crossed sometime in the 2040s, though every long-range forecast in this field has been wrong in both directions.
The physical obstacles are severe. At the storage densities of current hard drives, a yottaoctet would need something like a hundred billion units. Manufacturing them at present rates would take centuries, powering them would need the electrical output of many countries, and housing them would require a building programme without precedent. The number is not absurd, but it is far beyond the current industrial base.
Research into higher-density storage exists partly because of this ceiling. DNA data storage, which encodes information in synthetic genetic sequences, offers densities millions of times higher than magnetic media, and a yottaoctet of DNA would fit in a room rather than a continent. Reading and writing it remain slow and costly, but the density argument is what keeps the field funded.
The yottaoctet also appears in claims that turn out to be exaggerated. Reports that intelligence agencies were building yottaoctet-scale facilities circulated widely in the 2010s and were not supported by the construction, the power supply or the storage market. Any claim about a yottaoctet of anything can be checked against total world manufacturing, which is a useful discipline.
Between 1991 and 2022 the yotta prefix was the top of the metric ladder, which is why it was the natural unit for speculative claims. The addition of ronna and quetta in 2022 gave the system three more decimal orders above it, and it is telling that this was done partly because data quantities were approaching the old ceiling.
One yottaoctet equals 1,000 zettaoctets, 1,000,000,000,000,000 gigaoctets, 8 yottabits, or about 0.8272 yobioctets.