Two different "data" units, two different bases
Data storage sizes (kilobytes through terabytes) traditionally use a 1,024-based binary convention — 1 TB = 1,024 GB. Data transfer rates (kbps through Gbps) use a 1,000-based decimal convention instead — 1 Gbps = 1,000 Mbps. Both describe quantities of data, but scale by different factors at every step.
Where this becomes visible: a "1TB" hard drive
Using the binary convention, 1 TB works out to exactly 1,099,511,627,776 bytes. But many drive manufacturers label capacity using the decimal definition — exactly 1,000,000,000,000 bytes for "1 TB." Since an operating system reports the same physical drive using the binary convention, that manufacturer-labeled "1TB" drive shows up as roughly 931 "real" GB in Windows or macOS — not because storage is missing, but because two different definitions of "1TB" are being compared.
Why the two conventions never fully reconciled
Binary (1,024-based) prefixes fit naturally with how computers represent numbers internally, which is why operating systems adopted them early on. Decimal (1,000-based) prefixes match the standard metric system that manufacturers, and international standards bodies, prefer for marketing and specification purposes. Both conventions remain in active, legitimate use — neither one is simply wrong.
What this means practically
When a storage size and a transfer rate both appear in the same context — like estimating how long a file will take to download — converting each separately with its own correct convention (1,024-based for the file size, 1,000-based for the connection speed) avoids compounding two different kinds of unit confusion into one calculation.