>_sudo dojo
DashboardLearnCompete
Log inSign up

CompTIA A+ · Hardware

Storage Interfaces & RAID: Speed, Redundancy, and Resilience

Go deeper on how storage connects to the motherboard (SATA vs NVMe), the form factors that house them (M.2, U.2, 3.5-inch), and RAID arrays that trade speed and capacity for redundancy.

24 min read · Storage Interfaces & RAID

// CHECK YOUR KNOWLEDGE

You already know that SSDs are faster than HDDs, and that form factors like M.2 exist. This lesson goes deeper: the interfaces that connect storage to the motherboard (SATA and NVMe), the physical form factors that house them, and RAID arrays that let you combine multiple drives for redundancy or performance.

Storage Interfaces: SATA vs NVMe

The interface is the protocol and physical connection that a storage device uses to talk to the motherboard. Two standards dominate:

SATA (Serial ATA) is the older standard, used by both HDDs and some SSDs. SATA drives connect via a data cable and power cable to the motherboard and PSU. Speed is the main limitation: SATA maxes out around 550 MB/s on modern drives, regardless of whether it's an HDD or SSD.

NVMe (Non-Volatile Memory Express) is the modern standard for high-speed SSDs. NVMe drives connect directly to an M.2 slot on the motherboard with no cables needed. Speed depends on the PCIe generation:

  • PCIe 3.0 NVMe: ~3,500 MB/s
  • PCIe 4.0 NVMe: ~7,000 MB/s
  • PCIe 5.0 NVMe: ~14,000 MB/s

NVMe is much faster than SATA and is the standard for new systems. SATA persists mainly for HDDs and legacy SSDs; new SSD purchases are almost always NVMe.

▸ EXAM TIP

Exam tip: The A+ exam expects you to know that SATA maxes out around 550 MB/s, while NVMe is dramatically faster (3,500+ MB/s depending on PCIe generation). Understand that this interface difference is why an NVMe SSD feels much faster than a SATA SSD for boot time and large file transfers.

Form Factors: M.2 vs U.2 vs 3.5-inch

The form factor is the physical shape and size of the storage device. Different form factors use different interfaces:

M.2 — the compact stick-shaped format, used by most modern NVMe SSDs and some SATA SSDs. M.2 drives plug into an M.2 slot on the motherboard (no cables). They are named by width and length in millimeters — for example, 2280 means 22mm wide by 80mm long. Common sizes are 2230, 2242, 2260, and 2280. 2280 is the standard for desktop and laptop upgrades, while 2230 is common in handheld consoles (like Steam Deck and ROG Ally) and ultra-thin devices (like Surface). M.2 slots can be wired for SATA or NVMe (or both), depending on the motherboard. Always check the motherboard spec if using a SATA M.2 drive — they're slower than NVMe but cheaper.

U.2 — a larger form factor (looks like a 2.5-inch drive externally, but with a U.2 connector instead of SATA). U.2 connects via a dedicated U.2 cable to the motherboard and is primarily used by high-performance NVMe enterprise drives. You rarely see U.2 on consumer boards; it's more common in servers and workstations.

3.5-inch — the full-size desktop HDD form factor. Requires a SATA data cable and SATA power connector. Standard for large-capacity HDDs.

2.5-inch — used for laptop HDDs and SATA SSDs. Smaller than 3.5-inch. Can be used in desktops with an adapter bracket.

▸ COMPAT

Compatibility rule — form factors: M.2, U.2, and 3.5-inch are different physical shapes that use different connectors. An M.2 drive won't fit a 3.5-inch bay, and a 3.5-inch HDD won't fit an M.2 slot. Always check both the drive's form factor and the motherboard's available slots/bays.

eSATA and Hot-Swap

eSATA (external SATA) is an external variant of SATA used for external hard drives and enclosures. It uses the same protocol as internal SATA but is keyed differently to prevent accidental insertion into an internal SATA port. eSATA provides hot-swap capability — you can connect and disconnect drives without powering down the system.

Hot-swap means a drive can be added or removed while the system is running. This is useful for servers, backup systems, and any scenario where uptime matters. Not all drives support hot-swap (it depends on the controller and how the drive is mounted), but the SATA interface itself supports it.

RAID: Redundancy and Performance

RAID stands for Redundant Array of Independent Disks — a technique that combines multiple drives into a single logical unit for redundancy, performance, or both.

RAID 0 (Striping) — data is split across two or more drives, with each drive holding part of the data. If one drive fails, all data is lost. Minimum drives: 2. Benefit: maximum capacity and speed (if you have 2 × 1TB drives, you get ~2TB usable and roughly 2× throughput). Risk: zero redundancy — one failure = total data loss.

RAID 1 (Mirroring) — two drives hold identical copies of all data. If one fails, the other has a complete backup. Minimum drives: 2. Benefit: redundancy. Penalty: 50% capacity loss (2 × 1TB drives = ~1TB usable, because half the space is a mirror). Write speed is about the same as a single drive (the same data is written to both mirrors), but read speed can be faster — up to roughly double under ideal conditions — because a controller can read from both mirrored drives at once.

RAID 5 (Striping with Parity) — data is split across three or more drives, with one drive's worth of space devoted to parity (recovery information). If any one drive fails, the array can be rebuilt from the surviving drives. Minimum drives: 3. Benefit: good redundancy and capacity (3 × 1TB drives = ~2TB usable). If a drive fails, you can replace it and the array rebuilds. Risk: rebuild time is slow (can take hours or days for large drives), and if a second drive fails during rebuild, data is lost.

RAID 10 (1+0, striping of mirrors) — combines RAID 1 (mirroring) and RAID 0 (striping). Data is mirrored in pairs, then those pairs are striped. Minimum drives: 4. Benefit: high redundancy and speed. Penalty: 50% capacity loss (4 × 1TB drives = ~2TB usable). Can survive the failure of one drive per mirror pair.

▸ NOTE

RAID trade-offs summary:

  • RAID 0 — speed + capacity, but no redundancy (all eggs in one basket).
  • RAID 1 — redundancy, but 50% capacity loss. Write speed similar to a single drive, but read speed can be faster.
  • RAID 5 — balanced redundancy and capacity, but slow rebuild; use for read-heavy workloads.
  • RAID 10 — redundancy + speed, but 50% capacity loss; the "Goldilocks" choice for mixed workloads.

RAID Controller and Boot Considerations

RAID is managed by a RAID controller, which can be:

  • On the motherboard — built-in RAID support (common on most boards). Software-based, managed by the OS or firmware.
  • On a dedicated card — a PCIe expansion card with its own processor. Used in high-performance servers and workstations for better performance and features.

Boot considerations: The motherboard (BIOS/UEFI) must support booting from the RAID array. Most modern motherboards do, but you must configure the RAID volume before installing the OS. Once configured, the RAID array appears as a single drive to the operating system, and you boot from it like any other drive.

▸ WARNING

RAID is not a backup: A RAID array with redundancy (RAID 1, 5, 10) survives single-drive failure but is not a substitute for backups. Ransomware, accidental deletion, file corruption, and controller failure can destroy a RAID array. Always back up critical data separately.

Putting It Together

Storage architecture decisions depend on your needs:

  • Speed + capacity (no redundancy) — RAID 0 or a single fast NVMe SSD.
  • Redundancy — RAID 1 (simple), RAID 5 (better capacity), or RAID 10 (best of both).
  • Interface choice — NVMe for new systems (3–14× faster than SATA); SATA for budget or legacy systems.
  • Form factor — M.2 for desktops and laptops, 3.5-inch for large-capacity HDDs, U.2 for enterprise.

The check questions below test your understanding of interfaces, form factors, and RAID trade-offs.

Sign in to check your knowledge and earn XP. Sign in