Inside the lab follow MDrepairs on YouTube, TikTok & Instagram - visit our YouTube channel ↗

Data Recovery Guide

SSD Reports 8MB or 0MB Capacity? "Bricked Mode" Explained

Updated November 18, 2026
In this article

An SSD reporting 8MB, 32MB, or 0MB has lost normal capacity reporting or access. This may involve firmware, service metadata, the controller, or NAND. Different devices use different fallback states: Intel BAD_CTX and Phison SATAFIRM S11 should not be treated as one universal bug. Do not write to or reset the drive when you need its data.

An implausible capacity is a symptom, not a model-independent diagnosis. Similar reporting problems can occur on other storage devices too. Do not use secure erase, formatting, or a speculative firmware update to recover the contents.

What "Bricked Mode" Actually Means

Every SSD has a controller chip. The controller runs firmware that handles everything between the host computer and the NAND flash chips. It maintains the FTL (flash translation layer), which is the giant lookup table that maps logical block addresses the OS asks for to the physical NAND pages where the data actually lives.

When the controller boots, it reads the FTL from a protected area on the NAND. If that read fails, or the FTL is corrupted, or the controller hits an assertion in its own firmware code, it can't safely respond to read or write requests. The safest thing for the drive to do at that point is refuse to expose user data. So it falls into a safe mode where it reports a tiny "capacity" (usually 8MB for historical reasons) and waits for the manufacturer to send a recovery command.

The 8MB number is not magic. Different controllers use different fallback sizes:

  • Crucial M4 and MX series: 0MB or "no capacity" after the famous 5184-hour bug
  • Samsung 840 EVO/Pro and 850 EVO: 0MB or device disappears entirely after firmware faults
  • Intel 320, 520, 530, 600p: 8MB after a bad shutdown event (the original "SATAFIRM C2" bug)
  • OCZ Vertex 2/3: Drive vanishes, sometimes returns as 8MB after a long power cycle
  • SandForce-based drives (Corsair, ADATA, Patriot, Kingston, OCZ): 8MB common, sometimes 1GB or 6GB depending on revision

The Famous SSD Bugs That Made This A Thing

Bricked mode existed before any of these bugs, but a few high-profile cases brought it to mainstream attention. Knowing your drive's history helps figure out what's likely happening.

The Crucial M4 5184-Hour Bug

The Crucial M4 had a historical powered-on-hour firmware issue. That does not diagnose every M4 failure or imply one universal tiny-capacity recovery method. Consult the exact manufacturer's notice for a healthy drive; an unstable drive holding needed data should be assessed before maintenance.

The Samsung 840 EVO Read-Slowdown

Samsung 840 EVO read-performance issues and a drive reporting an incorrect capacity are different observations. A historical performance-restoration utility is not a general recovery method. Do not assume it applies to a failed SSD.

The Intel 8MB Bug (SATAFIRM C2)

Intel 320 BAD_CTX/8MB is a model-specific historical failure pattern. A manufacturer fix for future occurrences does not establish a data-preserving rescue of an already affected drive. Preserve it for diagnosis rather than attempting reset sequences.

OCZ Vertex 2 Disappearing Drives

OCZ Vertex 2 drives with SandForce SF-1200 controllers could brick after power events. The drives would disappear from BIOS entirely, then occasionally come back as 8MB after a 30-minute power-cycle hold. OCZ went out of business and got absorbed into Toshiba, leaving thousands of drives without firmware support.

Crucial MX500 Wear-Amplification

The Crucial MX500 has a known issue with high write amplification that wears NAND faster than expected. Drives running heavy database or VM workloads can wear out and brick within 2 to 3 years instead of the advertised 5+ year lifespan. We have a deeper guide on MX500 failures here.

How to Tell If Your Drive Is in Bricked Mode

You don't need lab gear. A working computer and 10 minutes is enough to confirm.

Step 1: Check what BIOS sees

Boot into your motherboard's BIOS or UEFI setup. Look at the storage device list. If your SSD shows up but the capacity is 8MB, 32MB, 0MB, or it lists a weird model name like "BAD_CTX" or "SATAFIRM S11" instead of the real product name, you're looking at bricked mode.

Step 2: Check what the OS sees

Boot from another drive (Windows install on a different SSD, or a Linux Live USB). Open Disk Management on Windows or run lsblk on Linux. If the drive appears with a tiny capacity, that confirms what BIOS reported. If the OS shows the right size but can't read partitions, you have a different problem (FTL corruption, not bricked mode).

Step 3: Run smartctl

A read of available health information can help diagnosis, but a particular error count or Available Spare value does not define a universal bricked mode. Identify the actual model and distinguish SATA SMART from NVMe health-log fields.

Step 4: Confirm it's not the cable or port

Try the drive in a different SATA port and with a different cable. If it still reports the wrong capacity, it's the drive. If it suddenly reports correctly, you had a flaky connection. Take that win and back up the drive immediately because the connection scare suggests other issues.

Important distinction: A drive that doesn't show up at all is not in bricked mode. It's just dead from the OS's perspective. Bricked mode means the drive responds but reports an obviously wrong tiny capacity. If your SSD isn't showing up at all, read this.

Why SSDs Fall Into This State

Five main triggers.

1. Power Loss During a Write

SSD writes are not atomic. The controller has to update the FTL after each write, and it does so in stages. If power dies between stages, the FTL ends up half-updated. On boot the controller checks the FTL, finds it inconsistent, and refuses to expose user data. Enterprise SSDs have power-loss protection capacitors that finish the write even after AC power dies. Consumer SSDs usually do not, which is why a UPS is worth the money on a desktop.

2. NAND Block Failure During Background Operations

SSD controllers run background tasks: garbage collection, wear leveling, read scrubbing. If one of these tasks hits a NAND block that's failed, and the controller can't fall back to a replacement block (because the spare pool is exhausted), the controller treats the situation as unrecoverable and bricks itself.

3. Failed Firmware Update

You hit "update firmware" in the manufacturer's tool. The update started, then dropped midway because the power blinked or the USB cable wiggled or the manufacturer tool crashed. The controller is now running a partial firmware that doesn't have all the modules it needs. It can boot only enough to report itself, not enough to access user data.

4. Firmware Bug Triggered at a Specific Condition

The Crucial M4 5184-hour bug is the textbook example. Some condition in the firmware (uptime, wear count, write count, temperature) triggers an unhandled code path. The controller panics and bricks itself. This is rare in 2026 compared to a decade ago because manufacturers got better at testing, but it still happens, especially on budget brands.

5. Voltage Spike or Static Damage

An ESD event (touching the drive's PCB with a charged finger) or a voltage spike from a bad PSU can damage the controller's PMIC (power management IC) or one of its data lines. The controller boots but can't reliably read from one or more NAND channels. To avoid returning bad data, it bricks itself.

What NOT to Do (Read This Carefully)

Bricked mode is the data recovery scenario where the wrong button click does the most damage. Here's the don't list. Print it if you have to.

Do NOT:
  • Run secure erase, sanitize, or a security reset only when deliberately discarding data. Depending on the device and command, these can erase storage or destroy encryption keys. Restoring the drive to use is not recovering its old files.
  • Format the drive or assign a new partition. Same problem. Format writes new structures over whatever's still readable.
  • Run a firmware update. Same problem, plus you can brick the controller harder. A firmware update on a bricked drive can change the controller's identity so even a lab can't get back in.
  • Run "fixers" you find on forums. Tools like MPTool, Phison utilities, SMI flashers are designed to flash firmware. They wipe the drive in the process. The forum post that says "fixed my 8MB Kingston!" left out that the user lost all their data.
  • Do not issue TRIM or deallocation as a repair. These commands mark logical data unneeded and can remove remaining access to it; they do not reconstruct a damaged translator.
  • Plug it into a USB enclosure if it was an internal SATA drive. Most USB-to-SATA bridges issue weird sequences of commands during enumeration. Some of those commands wake up the bricked controller and trigger more aggressive recovery attempts (which can wipe data).

The single rule in bricked mode: don't write anything to the drive. Don't run anything that might cause the drive to write to itself. Power it down. Get it to a lab.

Can You Fix It Yourself? Honest Answer

Sometimes. If your drive is a known model with a known fix (Crucial M4 5184, Intel 320 8MB), there's a sequence of power cycles and firmware steps that can sometimes unstick it. The trick is the order matters, and getting it wrong wipes the drive. If you're a hobbyist tinkerer with no important data on the drive, sure, try it. If you have a wedding's worth of photos on there, don't try.

The general (and not always successful) self-recovery for the famous bugs looks like this:

Power-cycle method (works on a fraction of bricked drives)

1. Remove the drive from the system entirely. 2. Disconnect it from any power. 3. Wait 30 minutes for capacitors to fully discharge. 4. Reconnect SATA data only, not power. 5. Connect SATA power. 6. Boot. Check BIOS. If the drive reports full capacity, immediately back up everything and replace the drive. This works on some Crucial M4, Intel 320, and OCZ drives but never on Samsung or modern Phison-based drives.

Heads up: Even the power-cycle method occasionally makes things worse. Some controllers attempt a self-recovery that finishes a wipe started before the brick. If your data is critical, don't try this at home. Send it to the lab in the state it's currently in.

Inside the Lab: Reading Past a Bricked Controller

When a bricked SSD arrives at MDrepairs, we have several options depending on the drive and what's wrong. This is roughly how the workflow goes.

Step 1: Read the Controller's State Without Letting It Run

For the exact controller and firmware combination, the lab checks whether a supported utility can provide technological access or virtual translation. PC-3000 SSD needs sufficiently functioning electronics and surviving information; ordinary imaging support does not establish advanced firmware recovery capability.

Step 2: Diagnose the Actual Problem

Based on what we see in service mode, we identify the cause. Common findings include corrupted FTL pages, exhausted spare blocks, dead NAND channels, or firmware module checksum failures. Each cause has a different repair path.

Step 3: Fix the Controller's Boot Path

A supported recovery utility may use a compatible temporary loader and surviving service metadata to reconstruct access. Factory formatting, clearing defect information, or reinitializing NAND can destroy data. Do not confuse those repair features with a data-preserving recovery procedure.

Step 4: Image the User Area

Once the controller is willing to expose user data, we image the entire drive sector by sector to a healthy target drive. We work from the image, not the original. The original goes back in its anti-static bag and sits as evidence. If the controller gets shaky during imaging, we can resume from where we left off.

Step 5: Mount and Extract

The image is then mounted as a virtual drive. If the file system is intact, we copy your files. If the file system is damaged (often the case on drives bricked mid-write), we run file system repair tools on the image, not on the original. More on our SSD recovery process here.

Why Chip-Off Is Not a Universal Fallback

Modern SSD recovery depends on a viable controller and encryption path. The lab assesses original-board repair, controlled reading, and firmware or virtual-translator access supported for the exact device. Independent raw-NAND chip-off cannot bypass encryption, and a lost controller path can make recovery impossible.

What a Raw NAND Dump Still Requires

An intact NAND package is not sufficient to recover an encrypted SSD. Rebuilding page order, ECC, or the flash translation layer does not supply missing keys. Preserve the original electronics; only a specifically supported method or repairable original read path can make extraction possible.

  • How the controller interleaved data across multiple NAND chips (channel/way layout)
  • How it ECC-corrected raw NAND reads
  • How it shuffled blocks for wear leveling (the FTL mapping)
  • What encryption the controller applied (modern drives encrypt by default)

The Encryption Problem

Internal encryption predates 2018 on some SSDs. Neither age nor absence of a password establishes plaintext NAND. If the necessary original encryption path or keys are irretrievable, independently reading chips does not recover usable files.

For an SSD with a failed controller, the first question is whether the original reading path can be restored. A same-model donor does not automatically decrypt the original NAND. Some legacy unencrypted designs support raw-NAND reconstruction, but that is not a general fallback for modern SATA or NVMe SSDs.

Older Drives Without Encryption

Only specific legacy unencrypted SSD designs with supported reading, error correction, scrambling, and translation are candidates for raw-NAND reconstruction. A cutoff such as 2014 is not a reliable way to identify them.

How Much Does Bricked SSD Recovery Cost?

Cost depends on what's wrong with the controller and whether the drive uses encryption. Rough brackets below.

Drive StateTypical RangeWhy
FTL corruption, controller responds to service tools$1,000 to $1,500FTL repair plus imaging
Bricked controller, donor firmware available$1,000 to $1,500Donor sourcing and firmware reload
Dead controller, donor swap on PCB$1,500 to $2,000+BGA rework, matched donor required
Original-board / supported controller recovery$1,500 to $2,000+Requires a viable original reading/encryption path; chip-off is not a general fallback
Original-board / supported controller recovery$1,500 to $2,000+Requires a viable original reading/encryption path; chip-off is not a general fallback

The $100 diagnostic is credited toward a successful recovery and is non-refundable once diagnostics start. Donor drives or replacement parts you approve are non-refundable once ordered, even if recovery fails. You approve the estimate before recovery work starts. Our full pricing page covers all data recovery scenarios.

Preventing This Next Time

SSDs will keep failing. Bricked mode will keep happening. What you can do is make sure when it happens, you have options.

Run a UPS on Important Machines

A $90 uninterruptible power supply on your desktop or workstation provides backup power during an outage. The UPS holds power long enough for the SSD's internal capacitors to finish current writes and for the OS to shut down cleanly. If you live in an area with even occasional brownouts, this is the cheapest insurance you'll buy.

Don't Run Firmware Updates Unless You Need To

Manufacturer dashboards prompt you to update firmware on a regular basis. Most of the time, the update fixes nothing that affects you and creates a small risk of bricking. Update only if you're hitting a specific bug the new firmware addresses. Crucial's support page and similar pages from other brands list known issues per firmware revision. Read before you click.

Keep a Backup Older Than 24 Hours

The window where things go wrong is right when you don't have a recent backup. A bricked SSD with a Time Machine or Backblaze backup from yesterday is annoying. A bricked SSD with no backup is a disaster. Pick a backup service, configure it once, leave it running. When is data recovery worth it vs starting over covers the calculus in detail.

Replace Aging Drives at 5 Years

Consumer SSDs are usually warranted for 3 to 5 years. If you have an SSD past 5 years old in a production machine, plan a replacement.

Avoid the Bargain Bin

SSD recovery cost and timing depend on original-board repair, supported firmware access, and the work needed to extract readable data. A dead encrypted controller can make recovery impossible rather than simply more expensive. Obtain a case-specific diagnosis and quote; chip-off is not a promised fallback.

FAQ

Why does my SSD say 8MB and not 0MB?

Different controllers use different fallback values. 8MB happens to be the safe-mode size on most Intel and SandForce-based controllers. Samsung often falls back to 0MB or simply disappears. The number doesn't tell you anything important about recoverability, just confirms you're in bricked mode.

Can I recover the drive with a firmware update?

Do not use firmware updates or reset sequences as recovery experiments on a failed SSD. A lab first assesses available access. Imaging can preserve readable data, but an SSD's physical firmware cannot be repaired merely by applying a firmware update to its disk-image file.

Is my data actually still on the drive?

Not necessarily. Data may remain physically present but still be inaccessible because of damaged NAND, missing translation information, unsupported firmware, or encryption. The reported capacity alone does not establish what can be recovered.

How long does bricked-mode recovery take?

SSD recovery cost and timing depend on original-board repair, supported firmware access, and the work needed to extract readable data. A dead encrypted controller can make recovery impossible rather than simply more expensive. Obtain a case-specific diagnosis and quote; chip-off is not a promised fallback.

What if my drive bricked during a firmware update?

A failed firmware update may leave some data intact, but it can also remove the available recovery path. Whether a compatible recovery loader or original-board repair helps depends on the exact device. A generic donor firmware module is not a guaranteed solution.

The drive worked yesterday. Could I have caused this?

Probably not. Bricked mode is usually caused by drive-internal events you have no control over. The exception is if you ran a "secure erase" or firmware update right before the issue. If you did, that's likely the cause. Tell us when you call so we know what state to expect.

Will SanDisk, Kingston, or PNY drives have this problem?

Yes. Any consumer SSD can brick. SanDisk SSD errors follow the same general pattern. Drives from PNY, ADATA, Patriot, and other budget brands often use SandForce or Phison controllers that have well-documented brick conditions.

Can I send the drive in an anti-static bag in a regular envelope?

Use a small box with padding. SSD PCBs are fragile when shipped loose. Anti-static bag plus bubble wrap plus a small box. Detailed shipping guide here. Our mail-in service page has the address and packing instructions.

Is there a TRIM-related risk during recovery?

TRIM or NVMe deallocation marks logical blocks as no longer needed. Their old contents may become inaccessible before garbage collection physically erases the NAND. Some supported SSDs allow specialist access to still-existing data, but erased or overwritten cells and lost required keys cannot be recovered. There is no universal time limit or guaranteed rescue window.

Will the data still be intact if the drive sat for a year?

Unpowered retention depends on NAND technology, wear, temperature, and the device's condition. There is no universal one-year safe-storage guarantee for a failed SSD. Keep it in stable, dry conditions and obtain assessment without unnecessary powering.

MDrepairs is an independent data recovery lab serving customers nationwide via mail-in service. We work with all major SSD brands. We don't sell drives or perform cosmetic repairs. Our specialty is getting data off devices that have failed.

Lost your data?
Let's assess your options.

Estimated turnaround: 4–6 weeks; timing may vary by case. Start with a free initial quote. The $100 lab diagnostic is credited on success; diagnostic and approved parts charges apply on failure.

Split the cost of your recovery with Affirm. Explore financing Subject to eligibility.

Devices recovered at our lab

These rows come from BenchKey, the ticket system we built for our own shop. Customer names, contact details and descriptions are left out.

Gloved hand holding a hard drive head stack at our Lincroft lab
A gloved hand holds a hard drive head stack at our Lincroft lab.
4.6 · 345 Google reviews$100 diagnostic, credited toward a successful recovery
Start your mail-in recovery