Conversion from 20 Petabits per second to Tebioctets per second

Information about the Petabit per second (Pbit/s)

The petabit per second is a unit of data transfer rate equal to a thousand terabits per second. Its symbol is Pbit/s. It marks the frontier of optical transmission research: the rate at which laboratories have carried data down a single strand of glass, and which no deployed system yet approaches.

The records were set by combining three techniques. Multi-core fibre puts several separate light-guiding paths inside one cladding, so that a single strand behaves like a bundle. Wavelength division multiplexing runs hundreds of distinct colours down each of those cores. Advanced modulation encodes many bits into each pulse. Multiplying the three together is what reaches a petabit per second.

Such demonstrations run over tens of kilometres of fibre in controlled conditions, not the thousands of kilometres a working sea cable must span. Distance is the hard part: signals attenuate, dispersion smears pulses together, and non-linear effects in the glass grow with power. Every kilometre added makes the same rate harder to sustain, which is why deployed cables sit two orders of magnitude below the laboratory record.

In octets a petabit per second is 125 teraoctets per second, which is roughly the storage of a hundred and twenty-five large hard drives moved every second. Nothing at either end of such a link could produce or absorb data at that rate; the figure describes the medium's capacity rather than any use of it.

At the network level the unit is used for aggregates. The total interconnect capacity of a very large data centre, the summed capacity of all the cables landing on a continent, and the peak traffic of the largest content networks are quoted in petabits per second. These are sums over thousands of links, not the rating of any one.

The trajectory is worth noting. Transoceanic capacity has grown by roughly a factor of ten every seven or eight years for several decades, driven almost entirely by better electronics and better coding rather than by more glass. If that continues, the petabit per second will describe a working cable within a generation.

One petabit per second equals 1,000 terabits per second, 125 teraoctets per second, or about 0.8882 pebibits per second.

Information about the Tebioctet per second (Tio/s)

The tebioctet per second is a unit of data transfer rate equal to 1,024 gibioctets per second, or two to the fortieth power octets per second. Its symbol is Tio/s. It is the binary counterpart of the teraoctet per second, and the two differ by 10 per cent.

Nothing outside a large machine moves data this quickly. The rate describes the memory bandwidth of an accelerator with stacked memory, the internal fabric of a high-end processor package, or the aggregate throughput of a parallel filesystem spread across thousands of drives. All of these are built from binary structures, and their totals are binary quantities divided by time.

High-performance computing is where the unit is written most often. A supercomputer's storage system is specified by how many tebioctets per second it can deliver to the compute nodes, because that number determines how quickly a simulation can save its state and resume. A machine that computes quickly but writes slowly spends its time waiting.

The ten per cent difference from the decimal unit is significant in that context. A filesystem procured to deliver 10 teraoctets per second and one delivering 10 tebioctets per second differ by a whole teraoctet per second, which in a facility of that size represents a substantial fraction of the hardware budget.

In bits a tebioctet per second is 8 tebibits per second, and in decimal terms about 1.1 teraoctets per second. Expressing the same rate four different ways is routine at this level, because the storage industry, the memory industry, the network industry and the standards bodies each prefer a different one.

For everyday comparison, a tebioctet per second would fill a large consumer hard drive in about twenty seconds. No external interface carries this; the figure describes movement between components inside a single system, where the wires are short and there are very many of them running in parallel.

Graphics processors have brought the rate within reach of a single component. A stack of high-bandwidth memory bonded directly to the processor die delivers well over a tebioctet per second to the chip that uses it, and a card carrying several such stacks passes a few. That bandwidth, rather than raw arithmetic speed, is what limits the training of large models: the arithmetic units sit idle unless the memory can keep them fed. The same reasoning explains why supercomputer designers spend as much effort on the paths between memory and processor as on the processors themselves, and why the rate is quoted in binary units when the memory it describes is addressed in powers of two.

One tebioctet per second equals 1,024 gibioctets per second, 1,099,511,627,776 octets per second, or about 1.100 teraoctets per second.