Windows 11 August Update Triggers NVMe Controller Failures; Phison Investigates

Microsoft’s August cumulative for Windows 11 has left a trail of disrupted installs and disappearing NVMe drives, and Phison — maker of widely used SSD controllers including the PS5012‑E12 — has confirmed it is investigating reports that the updates KB5063878 and KB5062660 can trigger controller-level failures during large sequential writes. The vendor says it is working with partners to identify affected controllers and deliver firmware or advisory remediation, but for now the practical guidance for users and administrators is blunt: back up critical data, avoid sustained large writes on suspect drives, and await coordinated firmware updates delivered through SSD vendors. eMicrosoft published a combined Servicing Stack Update (SSU) and Latest Cumulative Update (LCU) for Windows 11 (identified in community reporting as KB5063878, build 26100.4946). Within days, independent community testers and specialist outlets began reproducing a consistent failure profile: during sustained, sequential write workloads — commonly cited around the ~50 GB mark — some NVMe SSDs stop responding, vanish from File Explorer and Device Manager, and sometimes return with corrupted or inaccessible data. That operational fingerprint points to a controller hang or firmware-level lockup rather than a simple file‑system hiccup.
Those early reproductions over‑representnmilies (notably variants in the PS5012‑E12/E16 lineage), and a large number of affected models circulated in community lists — Corsair Force MP600 and several branded SanDisk, Kioxia, Maxio, and third‑party models were mentioned repeatedly. Not every drive with a listed controller failed in every rig, and several non‑Phison drives were also reported in isolated cases, which complicates single‑vendor attribution.

---

What Phison said — acknowledgement and next steps​

Phison issued a short, f made aware of the industry‑wide effects attributed to KB5063878 and KB5062660 and that some storage devices supported by Phison may have been impacted. The company emphasized it has engaged industry stakeholders, is reviewing controller families that may be involved, and will provide firmware updates and partner advisories as applicable. Phison’s message stresses partner‑centric remediation rather than direct public support, indicating firmware and vendor dashboards will be the distribution channel for any fixes.
This approach — centralized firmware delivery via SSD vendors rather than direct-to-consumer patches from Phison — is standard ause vendors must test controller firmware against branded drive hardware, factory configurations, vendor utilities, and customer‑facing firmware update mechanisms before broad distribution.

---

How the failure presents itself (symptoms and trigger profile)​

Core symptoms​

• The target SSD disappears from File Explorer, Device Manager, and Disk Management while a large write is in progress.
• SMART and controller telemetry become unreadable to diagnostic utilities and vendor dashboards after the event.
• A reboot sometimes restores cessarily file integrity; repeated heavy writes can reproduce the failure.

Typical trigger​

Community‑verified ron sustained sequential writes — often in the ballpark of ~50 GB or more — and elevated controller utilization percentages (c ~60%) as reliable triggers. Real‑world activities that match this profile include large game installs/patches, mass archive extraction, cloning, and video exports.

Why this is more than an annoyance​

When an SSD controller locks up, the OS often treats the device as if it were physically removed from the PCIe bus. That makes the drive invisible to the host OS and renders hs ineffective; in some reports the drive recovers after a power cycle, but in others data written during the failure window is corrupted or missing. The consequence is a genuine data‑integrity risk, not just a transient performance regression.

---

Technical analysis — plausible root causes​

Multiple independent analysts and community engineers have converged on two non‑exclusive hypotheses that explain how a Windows update can precipitate a controller‑level failure.

or NVMe driver regression​

A kernel or driver change in the update may alter command ordering, DMA timing, or buffer allocation behavior in ways that exercise a latent firmware bug. If the host issues NVMe commands in an order or cadence the controller firmware didn’t expect, the controller can stall or fail to respond. This would explain why drives from several vendors show similar symptoms under identical workloads.

2. Firmware edge cases exposed by HMB / DRAM‑less designs​

Many cost‑focused NVMe SSDs are DRAM‑less and rely on the Host Memory Buffer (HMB) feature to borrow host RAM for caching mapping tables. HMB behavior ties an SSD’s stability to host memory; subtle changes in HMB allocation or timing by the OS can expose race conditions or resource‑exhaustion bugs in controller firmware. Past Windows 11 24H2 rollouts previously exposed HMB‑related fragility in some drives, making this a plausible mechanism again.
Both mechanisms map cleanly to observed behavior: sustained sequential writes stress mapping tables, controller queues, and host memory interactions. Whether the primary fault resides in firmware (requiring a controller patch) or in host code (requiring a Microsoft hotfix or — remains subject to vendor telemetry and coordinated forensic analysis.

---

Which controllers and drives are implicated (what we know and what we don’t)​

Community collation and independent lab reproductions repeatedly flagged Phison controller families — including the PS5012‑E12 (E16 family), E21T and E31T variants — as being over‑represented among affected unitssurfaced frequently in early lists include the Corsair Force MP600 (Phison E12/E16 lineage), certain SanDisk Extreme Pro M.2 models, Kioxia Exceria Plus G4 SKUs, and a range of third‑party/white‑label drives.
Important caveats:

• These lists are community‑sourced and evolving; firmware revisions, motherboard UEFI, NVMe driver versions, and platform specifics materially affect whether a given drive will reproduce the failure. Treat model lists as investigative leads, not definitive recall inventories.
• Not every Phison‑;e model and controller behave normally in identical tests. That variability underscores the interaction nature of the failure — host behavior plus firmware state.

Given Phison’s confirmation that it is reviewing controllers and working with partners, expect firmware advisories to be issued thror than direct consumer channels from the controller maker.

---

Timeline and vendor/Microsoft response to date​

• Microsoft released the August cumulative update (KB5063878) on August 12, 2025. The public KB initially lisdevice issues.
• Within 48–72 hours, community testers and specialist outlets reproduced a consistent pattern of drives disappearing under sustained writes and began aggregating suspect models and controller fami acknowledged and mitigated a separate enterprise deployment regression (WSUS/SCCM error 0x80240069) via Known Issue Rollback (KIR). That fix addressed distribution but was distinct from the storage regreonfirmed it was investigating reported impacts to some storage devices and is working with partners to identify affected controller models and deliver firmware or advisory updates.

At the time of Phison’s r major SSD vendors were collecting telemetry and either issuing guidance or awaiting firm root‑cause determination before releasing firmware broadly. That staged, partner‑centric remediation path is deliberate — controller firmwagainst each vendor’s drive configuration and update mechanisms.

---

Practical mitigation and recommended actions​

The immediate priority for users and administrators is data protection are concise, conservative, and actionable.

For home users and enthusiasts​

• Back up critical data immediately to an independent device or cloud. A credible backup is the only reliable defense against drive‑level corruption.
• If KB5063878 or KB5062660 is not yet installed and your workstation relies on an NVMe SSD for critical work or you perform fconsider delaying installation until vendors and Microsoft publish remediation guidance.
• If you have already installed the KB and your system is stable, avoid performing large sustained writes (> ~50 GB) on suspect drives until firmware or Microsoft guidance is available.
• Check the SSD vendor’s management tool for firmware updates and advisories (Corsair iCUE, SanDisk ities, etc.) and only apply vendor‑recommended firmware after backing up.

For IT administrators and fleet managers​

• Inventory SSD models, controller families, and firmware versions across endpoints. Prioritize devices identified in earlDRAM‑less/HMB‑reliant SKUs.
• Use WSUS, MECM/SCCM, or your MDM to withhold KB5063878/KB5062660 from at‑risk machines until validated by vendor or Microsoft guidance. Microsoft’s KIRed to mitigate distribution regressions and should be monitored.
• Schedule large imaging, cloning, or bulk file distribution tasks on systems that are not yet patched, or use validated hardware ilures mid‑operation.
• If a drive becomes unreadable, preserve it: power off, capture system and NVMe logs, remove the drive to a forensic/quarantine host, and take an image if the data is critical. Vendors typicallmware version details for RMAs.

Registry and HMB tricks — caution advised​

Community workarounds used for previous HMB incidents (e.g., limiting HMB via registry tweaks) can sometimes reduce exposure but also degrade performance and are not a e fixes. Such mitigations should be treated as emergency stop‑gaps for advanced users with full backups and rollback plans, not broad production fixes.

---

Risks, tradeoffs​

• Data loss is the primary risk. The worst‑case scenarios include partially written or corrupted files and, in rare reports, volumes that remain inaccessible. Backups remain the single most important mitigation.
• **Staged vendor remediation isison will provide firmware updates to its SSD partners, and vendors will validate and publish those updates per SKU. This process reduces the risk of shipping an unsafe firmware but prolongs exposure for some end users.
• Attribution uncertainty increases operational friction. Because non‑Phison drives also appeared in some reproductions, blanket shutdowns or upgrade blocks may be organizations; targeted inventory and staged testing are the pragmatic paths forward.

---

Why this episode matters for the Windows and storage ecosystems​

Modern NVMe SSDs are complex systems combining controller firmware, DRAM or HMB-based metadata caching,e stack. When the OS makes seemingly small changes to buffering or memory allocation, it can expose latent firmware bugs that only manifest under heavy, sustained stress. This pattern — a host update triggering firmware edge cases — has precedent and demands better coordinati, silicon/controller makers, and SSD integrators during pre‑release stress testing. The August cumulative incident is a practical reminder that security and quality updates must be balanced with hardware compatibility verification for mass‑deployed components like SSDs.

---

#

• Phison and SSD vendors will continue collecting crash and failure reports and will issue firmware advisories or updates for affected controller families where a firmware fix is required. These updates will be distributed through SSD vendor channels and validated per SKU.
• Microsoft may add a Known Issues entry to the KB or deploy a targeted mitigation if root‑cause analysis shows the host OS behavior is a contributing factor. Historically, Microsoft has used both KIR and targeted fixes for similar problems.
• Administrators and power users should monitor vendor dashboards and communications for firmware packages and validation notes. When vendor firmware is released, test it in a controlled environment before mass deployment and ensure robust bacFinal analysis and practical takeaway

The available evidence shows a reproducible, workload‑dependent failure mode tied to recent Windows 11 cumulative updates that can, under sustained sequential writes, render some NVMe SSDs temporarily or permanently inaccessible; community labs and aggregators consin controller families among the high‑frequency suspects. Phison’s public acknowledgement confirms the company is investigating and working with partners to remediate affected controllers, which means firmware updates and vendor advisories are thution. Until those updates are validated and distributed, the safest course for users and IT teams is conservative: back up, pause heavy writes on at‑risk systems, inventory SSD firmware and controllers, and apply vendor‑recommended firmware only after valid reinforces a long-standing operational truth for modern PCs: OS updates change low‑level host behavior in ways that can cascade into hardware edge cases. The defense against those cascades is disciplined update management, vendor coordination, and, above all, current backups — the small cost that prevents a large, avoidable loss.

---

Conclusion: Treat the situation as an urgent compatibility warning rather than a confirmed universal recall. Expect Phison to publish targeted controller firmware updates through SSD vendors; until then, prioritize backups, delay heavy write workloads on suspect hardware, and follow vendor advisories for tested firmware or mitigation steps.

---

Source: Wccftech Phison Acknowledges SSD Failures On Windows 11 Due To KB5063878/KB5062660 Updates; Starts Investigations On Its Controllers To Resolve The Issue