Two conversions with different bases
Decimal SI storage units scale by 1,000: one terabyte (TB) is exactly 1,000 gigabytes (GB), and one TB is 10¹² bytes. Binary IEC units scale by 1,024: one tebibyte (TiB) is exactly 1,024 gibibytes (GiB), and one TiB is 2⁴⁰ bytes.
TB and TiB are not synonyms, just as GB and GiB are not synonyms. The uppercase B means byte; lowercase b normally means bit. Keep both the prefix and capitalization intact. A correct conversion begins by identifying whether the source is decimal or binary rather than applying 1,000 or 1,024 based on habit.
The names tera and giga belong to SI decimal prefixes; tebi and gibi are IEC binary prefixes. The inserted “bi” signals binary scaling. Using the proper names lets a reader recover the factor without guessing. A bare capacity number should never be converted until its source convention is known.
TB to GB and the reverse
Use GB = TB × 1,000. Reverse with TB = GB ÷ 1,000. Because both units belong to the same decimal family, the relationship is exact and no binary factor belongs in the calculation. A 2.5 TB decimal quantity is exactly 2,500 GB and exactly 2,500,000,000,000 bytes.
A positive TB value should yield a GB number one thousand times as large because GB is the smaller unit. Moving the decimal three places is a useful shortcut when the source and target are confirmed as decimal units. Preserve the unit label in spreadsheets and documentation so a later reader does not assume binary scaling.
Decimal examples remain exact even with fractional values: 0.125 TB equals exactly 125 GB and 125,000,000,000 bytes. A decimal fraction does not make the unit binary. Likewise, a storage product labeled 8 TB ordinarily requires its stated definition rather than an assumption based on the computer system where it will be installed.
TiB to GiB and the reverse
Use GiB = TiB × 1,024. Reverse with TiB = GiB ÷ 1,024. The factor follows from 2⁴⁰ ÷ 2³⁰ = 2¹⁰ = 1,024. A 2.5 TiB binary quantity is exactly 2,560 GiB and exactly 2,748,779,069,440 bytes.
Writing only “2.5 T” or “2,560 G” discards essential information. Explicit TiB and GiB labels communicate binary scaling. If software uses a nonstandard label, inspect its documentation or underlying byte count before deciding which calculation applies.
Binary scaling also produces exact integer relationships at prefix boundaries: 1,024 GiB equals 1 TiB, and 512 GiB equals 0.5 TiB. However, a quantity such as 1,000 GiB is not 1 TiB; it is 0.9765625 TiB. This distinction matters when setting thresholds or comparing a binary report with a round decimal target.
Three worked storage examples
Input: 4 TB. Formula: GB = TB × 1,000. Calculation: 4 × 1,000 = 4,000. Result: exactly 4,000 GB. Interpretation: both labels use decimal powers and represent 4,000,000,000,000 bytes. Reverse check: 4,000 ÷ 1,000 = 4 TB.
Input: 4 TiB. Formula: GiB = TiB × 1,024. Calculation: 4 × 1,024 = 4,096. Result: exactly 4,096 GiB. Interpretation: this is a binary-family conversion and should not be labeled 4,096 GB. Reverse check: 4,096 ÷ 1,024 = 4 TiB.
Input: 1 TB expressed in GiB. Formula: GiB = 1,000,000,000,000 ÷ 1,073,741,824. Calculation: = 931.322574615…. Result: about 931.32 GiB. Interpretation: crossing families requires bytes and explains why one decimal TB is less than one TiB.
Input: 0.5 TiB. Formula: GiB = TiB × 1,024. Calculation: 0.5 × 1,024 = 512. Result: exactly 512 GiB. Interpretation: binary halves remain exact because both units are powers of two. Reverse check: 512 ÷ 1,024 = 0.5 TiB. This fourth example also provides a quick benchmark for interpreting memory and image sizes.
Crossing between decimal and binary families
To convert TB to GiB, multiply TB by 10¹² bytes per TB and divide by 2³⁰ bytes per GiB. To convert TiB to GB, multiply TiB by 2⁴⁰ bytes per TiB and divide by 10⁹ bytes per GB. Do not simply replace 1,000 with 1,024 while leaving ambiguous labels.
The byte count is the stable bridge. It exposes which convention produced a displayed number and supports an independent check. At large scales, ordinary floating-point software may not represent every integer byte exactly, so systems that require exact accounting should use integer or arbitrary-precision arithmetic.
Confusing GB with GiB can affect quotas, backup planning, and capacity comparisons. Define units in reports and contracts rather than relying on organizational habit. If a threshold is specified in bytes, preserve that authoritative byte limit and convert only for display; do not replace it with a rounded human-readable value in validation logic.
Round-trip checks should use the unrounded internal value. Converting 1 TB to 931.32 GiB for display and then converting that rounded number back will not return exactly one trillion bytes. That is a presentation effect, not a failure of the definitions. Keep exact factors and sufficient numeric precision in calculations, and treat formatted text as output rather than authoritative input.
Raw, formatted, usable, and free capacity
A unit conversion describes the same byte quantity in another unit. It does not account for file-system metadata, partitions, recovery areas, firmware, redundancy, snapshots, reserved blocks, over-provisioning, or files already stored. Those factors can make usable or free space lower than a product's raw advertised capacity.
RAID and erasure-coding systems also use capacity for redundancy and may report values based on configuration. A converter cannot predict the result without drive counts, layout, failure tolerance, system rules, and definitions. Use product and operating-system documentation for a specific storage design.
Transfer rates use a time denominator and often bits rather than bytes. A network label such as Gb/s cannot be treated as stored GB. Estimating completion time also requires protocol overhead and real device performance. Capacity conversion is one input to such a model, not a complete forecast.
Virtual machines and cloud services may expose provisioned, allocated, consumed, billable, or available storage as separate measures. Thin provisioning can promise more logical capacity than is physically consumed at one moment. Unit normalization helps compare the numbers only after their operational meanings are identified; it cannot reconcile different service definitions.
Using the Data Storage Converter
Select TB, GB, TiB, or GiB explicitly in the Data Storage Converter. It converts through byte factors and also supports B, kB, MB, KiB, and MiB. It does not inspect hardware, calculate backup retention, estimate transfer time, measure compression, or determine available file-system space.
For a transparent comparison, record the original labeled value, its definition, the byte count when available, and the converted value with its full unit. Round only the display. This separates mathematical conversion from product-specific capacity, system use, and marketing presentation.
Large values should be displayed with separators and a clear unit, but data interchange should avoid locale-dependent punctuation. Store a numeric byte count or an explicit value-unit pair, then format for the reader. Never parse a human label such as “1.2 TB” as binary without documented rules, because the same text can otherwise lead to inconsistent quotas.