Microsoft’s August 12, 2025 cumulative update for Windows 11, KB5063878 (OS Build 26100.4946), is triggering a severe storage regression that causes certain NVMe SSDs to disappear from the system during sustained large sequential writes. Multiple independent community tests and tech outlets have confirmed the issue, which can lead to file corruption and unreadable SMART data, sparking urgent warnings for consumer and enterprise users alike.

The update, intended as the August quality and security rollup for Windows 11 24H2, includes security fixes and standard improvements. Official release notes make no mention of storage device failures. However, within days of the rollout, users began reporting that during file transfers of roughly 50 GB or more, affected SSDs would stop responding, vanish from Device Manager and Disk Management, and expose garbled controller data. A system reboot may temporarily restore the drive, but repeated heavy writes can reproduce the failure, often with corrupted files.

User Reports and Affected Hardware

Community-collated symptom patterns are strikingly consistent:
- The drive becomes unresponsive or “disappears” from Windows while a large, continuous file transfer is in progress.
- SMART attributes and controller information become unreadable by software tools.
- A reboot can restore visibility temporarily, but files written during the failure window risk corruption; some users report missing partitions after reboot.

The typical trigger is a sustained sequential write of approximately 50 GB or more. This workload is common during game updates, bulk media transfers, or disk cloning. Reproducibility varies by firmware revision, motherboard UEFI, and system configuration.

Affected drives often share a common thread: Phison controller families. Community lists include models such as:
- Corsair Force MP600 (Phison E16)
- SSDs built around Phison PS5012-E12 and E21T
- Kioxia Exceria Plus G4 (Phison-based SKUs)
- SanDisk Extreme Pro M.2 NVMe 3D
- Third-party brands like Fikwot FN955

These lists are not definitive, and not every Phison-based drive is affected. Isolated reports also point to a few HDDs and non-Phison NVMe drives. The key takeaway: DRAM-less designs that rely heavily on Host Memory Buffer (HMB) appear most vulnerable.

Technical Underpinnings: Why Large Writes Break Things

Large sequential writes stress multiple layers simultaneously: application buffers, OS page cache, kernel I/O scheduling, driver queues, and the SSD’s internal flash translation layer (FTL). A subtle regression in host-side NVMe driver timing or HMB allocation can push a controller into an edge case it previously dodged. When the controller locks up, the OS may treat the device as removed from the PCIe bus, hence the disappearance.

HMB is a critical factor. Many low-cost NVMe SSDs lack onboard DRAM and use a slice of system RAM for caching FTL metadata. Changes to HMB allocation size or timing—something past Windows 11 updates have done—can expose latent firmware bugs. Community analysis points to a host/driver interaction triggering a controller-level fault, but without vendor telemetry, the exact chain remains unconfirmed.

Official Responses: Silence from Microsoft and Drive Makers

As of publication, Microsoft’s KB5063878 support page lists no known storage issues. The company did acknowledge and fix a separate enterprise deployment problem: error 0x80240069 blocking installation via WSUS/SCCM, addressed in a servicing stack update. But the SSD disappearance has no official Knowledge Base entry.

Similarly, major SSD vendors have not issued consolidated bulletins attributing drive failures directly to KB5063878. Outlets like Igor’s Lab and Guru3D note the absence of official confirmation, urging caution before treating community model lists as recall notices. This gap leaves users in a precarious position: the evidence is strong but unofficially acknowledged.

Practical Steps for Consumers

If you rely on NVMe storage for critical data, immediate action is prudent:

  1. Back up everything now. Use an independent device or cloud service. This is your only reliable defense against drive-level corruption.
  2. If you haven’t installed KB5063878, and your workflow involves large sequential writes, delay the update until vendors or Microsoft issue clear guidance. Use Settings → Windows Update → Pause updates.
  3. If the update is already installed:
    - Avoid any sustained writes exceeding ~50 GB to NVMe drives, especially DRAM-less or Phison-based models.
    - Monitor drive health with vendor utilities (Corsair iCUE, SanDisk Dashboard, Kioxia Storage Utilities) or tools like CrystalDiskInfo and smartctl. Capture logs if anomalies appear.
    - If a drive vanishes during a transfer, stop all writes immediately. Do not repeatedly reboot; consider a forensic image if data is critical.
  4. Emergency mitigation (advanced users only): Some community members have limited HMB allocations via storport/stornvme registry parameters. This can act as a temporary stopgap but may reduce performance and carries risk. Back up the registry before attempting.

Guidance for IT Administrators

For enterprise fleets, the stakes are higher:

  • Inventory endpoints: Map drive models, controller families, and firmware revisions. Prioritize systems with DRAM-less NVMe drives and Phison controllers.
  • Hold KB5063878 deployment on at-risk hardware using WSUS/SCCM approval controls. Microsoft’s existing known-issue rollback for the 0x80240069 error demonstrates their capability to limit exposure; expect a similar mechanism if a broader storage issue is confirmed.
  • Coordinate firmware validation with SSD vendors before green-lighting the update. Test firmware updates in a lab environment; always back up before mass flashing.
  • Treat drive failures during imaging or provisioning as potential data-integrity events. Escalate to vendor support with logs, host traces, and a detailed reproduction recipe.

How to Verify KB5063878 Installation

  • Press Win + R, type winver, and check the build number. KB5063878 corresponds to OS Build 26100.4946.
  • Navigate to Settings → Windows Update → Update history to view installed updates.
  • On managed networks, check WSUS/SCCM deployment logs.

Recovering a Failed Drive

If an SSD has already stopped responding:

  • Immediately power off the machine and disconnect the drive to prevent further overwrites.
  • Capture diagnostics: vendor dashboard logs, smartctl outputs, and Windows Event Logs. This data is vital for vendor analysis.
  • For critical data, consult a professional recovery service for forensic imaging rather than attempting repeated repairs or reboots.
  • Contact the SSD vendor’s support channel with full details. If the failure is reproducible and widespread, they typically coordinate firmware advisories.

Weighing the Evidence

Strengths of the current findings:
- Multiple independent testers have reproduced the symptom set with methodical documentation.
- Several respected tech outlets (Notebookcheck, Guru3D, Igor’s Lab, WCCFTech) have aggregated reports, adding credibility.
- Microsoft’s rapid response to the WSUS deployment issue shows active servicing controls are in play.

Gaps and caveats:
- No vendor or Microsoft bulletin conclusively ties KB5063878 to a specific firmware bug. Multiple co-factors (UEFI version, driver states) may be involved.
- The causal chain—OS regression vs. latent firmware flaw—remains open. A fix could require an OS patch, firmware update, or both.
- The scale of incidents isn’t yet at a global-recall level; concentration likely depends on specific firmware and platform combinations.

Given these, the precautionary principle applies: minimize risk now and stay vigilant for official fixes.

Conclusion

The KB5063878 incident is a stark reminder that modern OS updates can interact with diverse hardware in unpredictable ways. Until Microsoft and drive manufacturers provide coordinated, tested remediation, users should treat large sequential writes on potentially affected NVMe SSDs as a high-risk operation. Back up religiously, defer the update on vulnerable systems, and watch official channels for firmware advisories or a servicing patch. The situation is fluid, but proactive defense can prevent data loss.