RAID Capacity Calculator

How each level spends capacity

The arithmetic is one subtraction per level:

• RAID 0 — no parity, so usable = drives × size. All 16 TB here, and no protection at all: one failure loses the array.
• RAID 1 — mirroring, so usable = size × (drives ÷ 2). Capacity buys redundancy one copy at a time.
• RAID 5 — one parity drive: (4 − 1) × 4 TB = 12 TB. The usual choice for four or more drives.
• RAID 6 — two parity drives, so 8 TB usable from 16 TB. Worth it on large drives where a rebuild can take a day.
• RAID 10 — mirrored stripes, 8 TB usable: the write performance of RAID 0 with the protection of RAID 1, at half the capacity.

Worked example: 4 × 4 TB in RAID 5

Raw = 4 × 4 = 16 TB.
Parity drives = 1, so data drives = 3.
Usable = 3 × 4 = 12 TB.
Efficiency = 12 ÷ 16 = 75%.

Move to RAID 6 with the same hardware and usable capacity falls to 8 TB while efficiency drops to 50% — the same disks, 4 TB less room, and protection against a second failure during a rebuild.

Where the calculator’s number stops

This is decimal terabytes: 1 TB = 1,000 GB, the way drive makers label them. Operating systems report TiB, so 12 TB of usable capacity shows up as about 10.9 TiB in Windows or macOS — nothing is missing, the units differ.

It also assumes whole drives of equal size. Mixed sizes are limited by the smallest member, and filesystem overhead, reserved blocks and snapshots take a further slice out of what you can store.

Mistakes worth avoiding

• Treating usable capacity as a backup. RAID protects against a dead drive, not a deleted file, a ransomware encryption or a controller failure that corrupts the set.
• Sizing an array on raw capacity. The 16 TB number is what you buy, not what you can write to.
• Choosing RAID 5 on very large drives. A multi-hour rebuild on a 20 TB disk has a real chance of a second, unrecoverable read error.
• Assuming any drive can be swapped in. Rebuilds want a disk at least as large as the smallest member.

Frequently Asked Questions