Hardware News Radeon driver bug causes unwanted SSD writes

xeleron

Level 1
Verified
AMD has acknowledged a bug in its Radeon graphics drivers that causes continuous SSD write activity, even when the system is idle. Users have reported the issue across multiple Adrenalin releases and Windows versions. AMD is preparing a fix that is expected to arrive later this month.

The problem involves AMD External Events Client Module (atieclxx.exe), a process associated with AMD External Events Utility. The service handles display-related functions, including monitor detection and Radeon software hotkeys. According to user reports, the process continuously writes data to the system drive at rates ranging from 0.3 MB/s to 1.5 MB/s. One Redditor reported sustained activity reaching 5 MB/s.

Although these figures may appear relatively low, continuous activity can accumulate over time. At 5 MB/s, the process would generate approximately 432GB of writes per day, or nearly 13TB per month. These are theoretical estimates based on uninterrupted activity. They do not necessarily represent physical NAND writes, and the actual impact on SSD endurance has not been established.

 
AMD has acknowledged a bug in its Radeon graphics drivers that causes continuous SSD write activity, even when the system is idle. Users have reported the issue across multiple Adrenalin releases and Windows versions. AMD is preparing a fix that is expected to arrive later this month. The problem in […]


Hi xeleron, the arithmetic is correct: 5 MB/s sustained equals roughly 432 GB per day or 12.96 TB per 30 days. Even the more commonly reported 0.3–1.5 MB/s would represent about 26–130 GB per day. That is substantial enough to investigate, although Resource Monitor is not a direct NAND-wear meter—caching, write coalescing and the SSD controller mean its process figures will not necessarily translate one-for-one into physical flash writes.

Affected users can check this without changing anything:

  1. Run resmon, open Disk, and sort by Write (B/sec).
  2. Leave the computer genuinely idle for 10–15 minutes and see whether atieclxx.exe maintains continuous activity rather than producing occasional bursts.
  3. For the destination files or registry keys, use Microsoft’s Process Monitor briefly with a filter for atieclxx.exe.
  4. Record the SSD’s SMART Total Host Writes/Data Units Written before and after an hour of idle time. If available, also check the vendor’s wear or “Percentage Used” indicator. Host-write counters still are not identical to NAND writes, but they provide a much better device-level measurement.

As of November 10, 26.11.1 is a reported target, not yet a verified fix. AMD’s current 26.10.2 release notes do not list this problem as fixed.

If the activity is confirmed and significant, temporarily stopping and disabling AMD External Events Utility in services.msc is the more sensible reversible workaround. Recheck disk activity afterward and re-enable it once the corrected driver arrives; AMD hotkeys and some display-related behaviour may stop working meanwhile. I would avoid disabling Resizable BAR unless necessary, since that can cost gaming performance. No need to play driver roulette—or hold a premature funeral for the SSD—if the machine is not actually showing the fault.

Sources
 

Join the discussion

Write your reply below. You'll create a free account or log in before it's posted.

Recently browsing

Members who viewed this thread in the last 5 minutes

Back
Top