Numbers, units, and bits
Compare GB with GiB and Mbps with MB/s
Tools to use
Capacity figures can disagree even when both are described as 'giga'. Likewise, a network advertisement and a download window may use bits and bytes respectively. Missing one character in a unit can substantially change a storage or transfer estimate.
The data-units tool lets you compare values after converting to bytes or bits per second. These experiments do not diagnose a defective drive or connection. Their purpose is to separate differences explained by notation from differences requiring investigation of actual storage use or network conditions.
Convert 1 GB and 1 GiB back to bytes
Select capacity, enter 1, and convert GB to byte: the result is 1,000,000,000 bytes. Keeping the input at 1 but changing GB to GiB yields 1,073,741,824 bytes. A unit change changes the quantity represented by that same number; it is not merely a different label.
1 GB = 1000 × 1000 × 1000 byte
1 GiB = 1024 × 1024 × 1024 byte
1 GiB - 1 GB = 73,741,824 byteGB uses powers of 1000 and GiB uses powers of 1024. The gap accumulates at larger scales. Treating them as interchangeable can distort a capacity plan as the number of devices grows. Keep units in table headings and check the convention used by each product or management screen.
Explain the roughly 931 shown for 1 TB
Convert 1 TB to GiB to obtain approximately 931.3225746154785 GiB. This divides the original 1,000,000,000,000 bytes by the bytes in one GiB. The difference so far is entirely a change in the number used to describe the same quantity, not a loss of bytes caused by conversion.
Do not use this conversion to explain all available-space differences. Filesystem metadata, recovery partitions, and existing files are separate factors. Distinguish advertised capacity, the operating system's total capacity, and current free space. A figure near 931 alone proves neither normal operation nor a fault; first compare equivalent measurements.
100 Mbps and 12.5 MB/s describe the same rate
Select rate and convert 100 Mbps to MB/s: the result is 12.5. One byte is eight bits, so divide by eight. The tool may label megabytes per second as MBps. The lowercase b in Mbps and uppercase B in MB/s denote different units; comparing the numbers alone would suggest an incorrect eightfold slowdown.
Changing the output to MiB/s gives approximately 11.920928955078125. Now both the bit–byte conversion and the decimal–binary scale matter. When rates are close but not identical, check for the i as well as capitalization. Report 12.5 MB/s rather than '12.5 mega' so the time basis is clear.
Conversion is not a throughput measurement
The conversion restates the same rate; it measures neither efficiency, congestion, storage performance, nor peer limits. A download at 10 MB/s does not make the entire difference from 12.5 MB/s a network loss. You must also align what was measured and over which part of the transfer.
For duration planning, use transfer time with an explicit efficiency assumption. Comparing 100% and 80% makes the assumption visible, but 80% is not a measured constant. If the entered speed is already measured throughput, applying the same loss again through efficiency double-counts it. Define what the speed represents first.
Why GB cannot be converted directly into Mbps
Capacity is an amount; rate is an amount per second. Converting 1 GB into Mbps requires a duration. Sending 1 GB in 80 seconds gives a theoretical 1,000,000,000 × 8 ÷ 80 = 100,000,000 bit/s, or 100 Mbps. Eight seconds requires ten times the rate. The time information cannot simply be omitted.
Separating capacity and rate prevents a dimensionally invalid operation from looking like a legitimate conversion. Use this tool to align quantities in the same dimension, and the transfer-time tool to ask what speed a deadline requires. Choosing the question first clarifies which inputs are missing.
Copy the scale along with the number
Pasting a GiB number into a column headed GB changes its interpreted quantity. Keep the output unit, original unit, and, where useful, the conversion formula. If rounding for presentation, retain the unrounded value for subsequent calculations rather than repeatedly converting a displayed 931.3 and accumulating rounding differences.
Try a round trip: convert 1 GiB to bytes, then convert those bytes back to GiB; the result should be 1. If it differs, check for intermediate rounding or mixed capacity and rate units. Establishing what unit conversion explains helps avoid unnecessary fault investigations or inflated capacity estimates.
When source and destination sizes look different
If one screen shows 1GB and another about 0.93GiB, did the transfer lose data? Start with the byte count. The same 1,000,000,000 bytes can be displayed using decimal or binary units without changing the content. A display rounded to two decimal places cannot recover the exact original byte count, however. Use the exact size when available, rather than converting an already rounded label back and treating it as evidence of loss.
If exact byte counts still differ, inspect what is being measured: a compressed archive versus extracted files, logical file size versus allocated disk space, or text whose line endings changed. Unit conversion does not diagnose these causes. Compare the same kind of size first. Equal lengths also do not establish equal content; checking a transfer's content requires a separate comparison, such as hashes against a trusted original.
An estimate such as '10,000,000,000 input bytes, before compression or replication' is more reusable than 'needs 10GB.' Retention days, backup generations or transfer efficiency are additional conditions, not unit conversions. Establish one byte count first and then apply the relevant model, so you can explain at which stage a size or cost increased.