forgive my ignorance, there must be a really simple answer to this, but im unaware of it, also, tried google-ing for it, but didnt know what to look for exactly, and didnt get much good answers either.

my question is, since 2^32 is around 4Gb, so max usable RAM on 32bit pc is 4GB.

then in what way HDDs with hundreds of GBs are happily supported by the same os...

regards
somjit.

edit: i guess i found my answer here

Dani AI

Generated

Short answer: the “32‑bit” number refers to CPU/virtual memory addressing, not to disk addressing. A 32‑bit process (or OS) has about 2^32 bytes of virtual address space (~4 GiB), which limits how much RAM a single process can directly map — it does not mean disks or files must be smaller than 4 GB. (en.wikipedia.org)

Storage uses different layers. Drives are addressed with Logical Block Addressing (LBA) at the hardware/firmware level (ATA introduced 48‑bit LBA in ATA‑6, which supports enormously large drives), and the partition table format (MBR vs GPT) and storage drivers present logical sectors to the OS. MBR uses 32‑bit sector counts and so effectively caps a single MBR partition at about 2 TiB (with 512‑byte sectors); GPT removes that limit but requires firmware/OS support to use as a boot disk. (This is why a 500 GB drive is trivially usable on a 32‑bit system.) (en.wikipedia.org)

File systems matter too. FAT32 stores file lengths in 32‑bit fields so a single file is limited to 4,294,967,295 bytes (~4 GB − 1). Modern file systems (NTFS, exFAT, ReFS) use larger offsets and support multi‑TB volumes and multi‑GB files; Windows exposes 64‑bit file offsets (SetFilePointerEx / LARGE_INTEGER), so 32‑bit applications can work with large files as long as the file system and drivers support them. (en.wikipedia.org)

Practical checks/tips (no fluff): use Disk Management or run diskpart and list disk to see whether a disk is MBR or GPT (GPT disks show an asterisk in the GPT column). If an MBR disk is larger than ~2 TB only the first ~2 TB is addressable; to use the whole drive the disk must be GPT and the platform must support GPT/UEFI for booting. Always back up before converting/initializing a disk, and consider spreading important data across multiple drives or using RAID/backups as suggested.

Recommended Answers

All 4 Replies

The maximum size of a hard drive partition is not determined by the operating system (32 bit or 64 bit). It is determined by the file system. Thus, NTFS can manage a larger partition than FAT or FAT32. Whether the medium is solid state memory or magnetic platters is irrelevant.

To enlarge...
The ATA-6 /Ultra DMA100 standard provides for a 48 bit LBA address space for sectors [Ultra DMA133 is just faster], so that is the current hardware limit, I believe. 2^48 * 512B sector size = 144 * 10^15, or 128 * 2^50. That is 128PB. 2^50 is a PiB.
NTFS itself uses 32 bit addressing internally, although it is capable of 64 bits; NTFS uses 4KB clusters as address blocks by default, so theoretically you have a partition size maximum of 16TB. 2^32 * 4096 = 17.6 * 10^ 12, or 16 * 2^40. 2^40 is a Tib.
But if you forgo file compression capability you can set a cluster size up to 64KB with the format command.... and the limit for NTFS is then 256TB. Who needs file compression on a volume like that?

Phew. Numbers.

Why would one want to keep so much data on a single drive? The ball bearings must be made with something really special in these gigantic GB and TB HDD for folk to have so much faith in them.
Call me old fashioned, but I prefer not to keep all my eggs (data) in the one basket(HDD) and even though small capacity HDD are becoming like hens teeth, any chance I get to buy them, I do.

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.