Question about advertised vs. actual storage sizes (SSDs, HDDs, flash drives, etc)
I'm aware that size can be measured in either decimal or binary, and that you will get two different numbers. Same size just measured differently. For instance, a 300 GB hard drive will show as both 300 billion bytes and 279 GB (or GiB, measured using binary). Both numbers correct, just decimal vs. binary.
Since computers measure using the binary system, whereas manufacturers use decimal, is the drive in this example giving you the full 300 billion bytes of space? It has always been my understanding that it utilizes the full 300, though there is a lot of conflicting info out there on the subject.
Thanks
279 * 1.024 * 1.024 * 1.024 = 299.574
With rounding errors, this is about as close to 300 GB (billion bytes) as you're going to get. So yes, a drive advertised as 300 GB, really does have 300 billion bytes available to the PC. You can confirm this by looking at the number of sectors on the drive - almost all drives advertise themselves as having 512 byte sectors (whether or not they actually do). For example, the WD Caviar green 3 TB makes 5,860,533,168 sectors available to the PC (5,860,533,168 * 512 = 3,000,592,982,016 bytes available).
However, in practice, not all of the capacity is available for files. Most OSs insist that the drive have 'partitions' marked onto it, and partitions have overhead. To store files, you need a 'file system' which indexes the files, organises them into folders, works out how to organise the file data on the disk, protects or restricts files, etc. File systems have more overhead - so the actual capacity the OS will make available for files will be less than advertised (but how much less will vary depending on the OS, file system, etc.)
Of course, if you decided to use your own custom partiton and formatting technique, you would have 300 billion+ bytes of functioning storage available to do with as you want.
The discrepancy in measurement units is about 7% for GB (1.024 * 1.024 * 1.024) , and about 10% for TB (1.024 * 1.024 * 1.024 * 1.024). Calculating the discrepancy for PB/PiB is left as an exercise for the reader.
Historically, the discrepancy is a lack of agreement between different groups of engineers as to the appropriateness of approximations and where the approximations are used. Historically, the binary units have always been used for RAM. However, they have almost never been used for anything else. Decimal has always been used for network/modem/serial speed, CPU speed and tape storage. Bulk storage was originally tape, and tape was specified in 'bytes (or bits) per inch' and the tape was measured in feet - so a 10 Megabyte tape would hold 10,000,000 bytes. Floppy disks were a development of tape and were often specified in decimal, but there was some variability, with a number of types of disk being specified using binary units. By and large, at this stage, most data storage systems were in the kB range - so a 2% discrepancy was barely noticable, so there was little impetus for programmers to standardise on units.
Hard drives were usually decimal, but again there was some variability with some using binary, and others using some weirdo hybrid schemes (e.g. defining 1 kB as 1024 bytes and 1 MB as 1000 kB = 1024000 bytes!!), particularly in the early days. However, once into the MB range, the discrepancy between units grew to 5% and people started noticing. The GB range made things worse.
By this stage, considerable established software had been developed using the binary unit scheme, and this convention was very powerful - Microsoft continue to use it, even in Windows 7. Interestingly, apple don't - they use decimal units.