NTFS vs ext4 vs APFS: Understanding File Systems Compared

Understanding file systems: NTFS, ext4 and APFS compared

You plug a USB drive into a Mac, copy a folder, then move that same drive to a Windows laptop. Nothing opens. The drive shows up, sure, but the files are either invisible or locked in "read-only." That moment right there is where most people learn that file systems exist at all. Not in a textbook, not in a class, but in the space of five frustrating minutes at a desk.

I've hit that wall more times than I'd like to admit. I once formatted a 2TB drive on a Mac, loaded it with camera footage, and handed it to an editor who worked on Windows. He called me twenty minutes later, mildly annoyed. Everything worked eventually, but only after we reformatted and copied the files again. NTFS, ext4 and APFS behave differently, and those differences have real consequences the moment you move data between machines.

So here's what actually matters: how these three file systems are built, where each one shines, and how to pick the right one without wasting an afternoon re-copying files.

What a file system actually does (and why you should care)

A file system is the layer that decides where each byte of your data physically lives on a disk, what it's called, who can touch it, and what happens when the power cuts out mid-write. That's the whole job, stripped of jargon.

Two mechanisms matter more than the rest.

Journaling

Before writing actual data, the system logs its intent in a small reserved area. If the machine dies halfway through, the log tells the system what was in progress, and it can roll back or finish the job. Without journaling, an unexpected shutdown can leave a drive in a state where the directory tree doesn't match the actual data blocks. Files become ghosts: listed but unreadable.

NTFS, ext4 and APFS all journal. That's the baseline in modern systems, not a feature worth bragging about.

Metadata and structure

Each system stores metadata differently, and that's where the real divergence starts. NTFS keeps a master file table. ext4 uses inodes with extents. APFS leans on a container that holds multiple volumes sharing the same free space pool.

I'll be honest: when I first read about APFS containers, I didn't get it. It sounded like marketing. Then I set up a Mac with three "separate" volumes and watched them dynamically borrow space from each other. That is the point. You stop guessing how to split a drive.

• NTFS — master file table, per-file permissions, native encryption
• ext4 — inode-based, extent-mapped, mature and predictable
• APFS — copy-on-write, snapshots, space-sharing containers

NTFS: the Windows standard

NTFS has shipped with Windows since the early 2000s and has barely needed a replacement. Microsoft's alternative, ReFS, exists mainly for servers and storage arrays, not for the laptop on your desk.

Where NTFS wins

It handles enormous volumes without complaint, supports files far larger than most people will ever create, and gives you file-level permissions that integrate with Windows accounts. Encryption is available natively through BitLocker on the volume side, and EFS at the file level. For a company running Windows machines, that combination is hard to beat.

Reliability is solid too. I've recovered data from an NTFS drive that took a hard power loss during a large copy operation. The journal did its job. The directory tree was intact, and only the file being written was damaged.

Where NTFS gets annoying

Outside Windows, it's a guest. macOS reads NTFS out of the box but refuses to write without third-party tools. Linux can mount it read-write through the kernel's ntfs3 driver, and it works well in my experience, but it's not the native environment. If your workflow crosses operating systems regularly, NTFS becomes a constant negotiation.

The permission model is also heavier than what a personal drive needs. Inheritance rules, access control lists, ownership — useful in an office, overkill for a drive holding movies.

ext4: the Linux workhorse

ext4 is the fourth generation of the extended file system, and it's been the default on most Linux distributions for well over a decade. Not because it's the most advanced option — Btrfs and ZFS offer more — but because it's boring in the best sense. It does its job and rarely surprises you.

What makes ext4 dependable

Extents replaced the old block-mapping scheme, which cut fragmentation and sped up large-file operations considerably. Delayed allocation lets the system batch writes intelligently. Directory sizes in the hundreds of thousands of files are handled without the performance cliff that older systems hit.

On a personal note: I ran a small file server on ext4 for years, holding a few million small files. It never once gave me a corruption scare. When I migrated the same data to a more exotic system to try the newer features, I hit two separate issues in the first month. I moved back to ext4 and stopped experimenting. Sometimes the boring answer is the right one.

The limits worth knowing

No native snapshots. No built-in checksumming of data blocks. No transparent compression. If you want those, you're looking at Btrfs or ZFS, both of which have their own trade-offs around maturity and complexity.

Windows can't read ext4 natively at all. There are third-party drivers that add support, but I wouldn't trust them with anything important. Linux and ext4 go together; everything else is a workaround.

APFS: Apple's rethinking of the file system

Apple introduced APFS around 2017, replacing the long-serving HFS+. It wasn't a minor update. The design is fundamentally different from both NTFS and ext4.

Copy-on-write and snapshots

When you modify a file in APFS, the system writes the new version to a fresh location rather than overwriting the old blocks. The original stays intact until the change is committed. This makes snapshots nearly instant and extremely cheap in terms of disk space, because unchanged data isn't duplicated.

Time Machine on macOS uses this heavily. Backups that used to take an hour now finish in minutes, because only the changed blocks get written. That's not a small improvement — it changes how often you're willing to back up.

Containers instead of partitions

A single APFS container can hold several volumes that all share the same free space. You don't predetermine that volume A gets 500GB and volume B gets 300GB. They grow and shrink as needed, within the container's total.

This is genuinely useful and genuinely confusing the first time you see it in Disk Utility. You wonder why your "partition" sizes don't add up to the disk. They're not partitions.

The catch

APFS is Apple-only. Windows won't touch it. Linux has read-only support at best in most setups, and even that depends on the distribution. If you format an external drive as APFS, you've committed to the Apple ecosystem for that drive's lifetime.

Encryption is also on by default on modern Macs. Great for security. Terrible for recovery if you lose the keys, because the encrypted metadata becomes unreadable and data-recovery tools have far less to work with.

NTFS, ext4 and APFS side by side

Feature	NTFS	ext4	APFS
Native OS	Windows	Linux	macOS
Journaling	Yes	Yes	Yes (copy-on-write)
Snapshots	Via shadow copies	No (needs LVM)	Yes, built in
Native encryption	BitLocker, EFS	Via LUKS	Yes, default
Read/write on Windows	Full	Third-party only	No
Read/write on Linux	Via ntfs3	Full	Read-only (limited)
Read/write on macOS	Read-only	Third-party only	Full
Best for	Windows internal drives	Linux servers and desktops	Mac internal drives

Notice the last four rows. Cross-platform access is where all three systems fail, in different ways. That's not a bug; it's a design decision. Each vendor optimizes for its own ecosystem, and interoperability is an afterthought.

Performance: what actually changes between these systems

Benchmarks comparing APFS, NTFS and ext4 appear regularly online, and the honest summary is that differences are smaller than most people expect on modern hardware. Sequential throughput — copying one large video file — lands within a narrow band across all three. The drive's physical interface usually matters more than the file system on top of it.

Random access patterns are where things diverge. ext4 with extents handles many small files efficiently. APFS copy-on-write adds overhead on heavy random writes but shines on snapshot-heavy workloads. NTFS sits somewhere in the middle, with strong performance on large volumes.

Here's the part benchmarks never capture: your workload decides. A video editor copying 50GB files won't notice the differences. A developer running a database with thousands of tiny random writes will. I've seen a build process speed up meaningfully just by moving the working directory to a different file system on the same physical disk.

Which file system is best for external hard drives?

For a drive that only ever connects to one type of machine, use that machine's native format. NTFS for Windows-only, APFS for Mac-only, ext4 for Linux-only. It's the simplest answer and the right one.

The moment a drive has to move between operating systems, though, the picture changes. None of the three big systems handles that cleanly, so exFAT becomes the practical default. It's supported read-write by Windows, macOS and most Linux distributions without extra software.

The trade-off is real, though. exFAT has no journaling, no permissions, and no encryption. It's designed for removable media, not for storing your only copy of something important. I keep an exFAT drive for moving files between machines and never treat it as a backup. If it dies, I lose the transit copy — not the source.

Frequently asked questions

Is ext4 better than NTFS?

Neither is universally better. ext4 tends to handle large numbers of small files efficiently and is the natural choice on Linux. NTFS offers richer permission and encryption features geared toward Windows environments. If you're choosing between them for a specific machine, pick based on the operating system that machine runs, not on abstract benchmarks.

Which is better, APFS or NTFS?

They're built for different worlds and don't compete directly. APFS is better for Macs because of snapshots, containers and default encryption. NTFS is better for Windows because of native integration and permission management. If you need a drive to work with both, neither is the right answer — that's what exFAT is for.

What are the four types of file systems?

There's no single official list, but the distinctions usually come down to how data is organized and where it's used. The broad categories are: disk file systems (NTFS, ext4, APFS, exFAT), network file systems (NFS, SMB), distributed or clustered systems (used across multiple machines), and special-purpose systems for things like flash storage or databases. Within disk file systems alone you'll find many more than four named options.

The format choice nobody tells you about

Most guides treat file systems as a technical detail. In practice, the decision you'll actually face is narrower: internal drives should match the operating system, and external drives should be formatted for whatever machines will touch them. Get that right and you'll rarely think about file systems again.

What surprises people is how often the "best" answer is the least exciting one. ext4 isn't the fastest or the most feature-rich Linux option. NTFS isn't the newest Microsoft design. APFS isn't perfect. But each has spent years proving itself on millions of machines, and that track record carries more weight than any benchmark chart.

The next time a drive refuses to open on a different computer, you'll know exactly why. It won't make the drive work any faster — but at least you'll understand what went wrong, and what to format it as next time.