AivexaNewsSearch
AI news for builders and product teamsChecked every hour

AMD confirms GPU driver bug is wearing out your SSD – how to stop it

Collected Nov 9, 2026

AMD has acknowledged that a component shipped alongside its Radeon graphics drivers is repeatedly writing data to users' solid-state drives, accelerating wear. The culprit is the AMD External Events Utility (EEU), which the company installs together with its CPU and GPU software. A company representative posting under the name AMD_Vik said on Reddit that the problem will hopefully be fixed in the next update, leaving users without a firm release date in the meantime.

Reports trace back to July, when a user on AMD's official PC user forum noticed that the External Events Client Module and EEU were writing to their SSD constantly after installing a Radeon RX 9060 XT card. Comments added to that thread this month indicate the behavior has continued through several bug fixes and patch releases. Two users there put numbers on it: sustained write activity ranging from roughly 0.3 MB/s to as much as 1.5 MB/s.

This week a Reddit user named darklinkpower posted that the AMD Adrenalin Edition desktop application also carried the bug, and that it had consumed up to 1 percent of their drive health per month over about 16 months. On its own that sounds modest, but SSD endurance is finite and irreversible: flash memory cells tolerate a limited number of program-and-erase cycles, and a background process that writes continuously burns through that budget even when the machine is idle. A drive that would normally sit quiet for hours ends up under constant small-write load.

The recommended workaround addresses the service directly. Open the Run dialog with Windows key + R, type services.msc and press Enter, find AMD External Events Utility in the list, open its properties, stop the running process and set its startup type to Disabled. Among users who have reported the problem, this appears to be the most reliable remedy. It comes with costs: the EEU is involved in display-related features, so turning it off can remove AMD FreeSync support, and can introduce display glitches when monitors are plugged or unplugged while the PC is powered on.

A second, more experimental option involves Resizable BAR, which AMD markets as Smart Access Memory. The technology lets the CPU address the GPU's full memory rather than a small window of it, typically for a performance gain in games. According to the report, the EEU bug can interact with this arrangement so that the utility and the GPU consume whatever memory they can find. Inside the Adrenalin Edition app, users can click the gear icon, open Settings, and check whether Smart Access Memory shows as Enabled, Disabled or Not Available; if it is on, toggling it off is the suggested step. Because this can disturb CPU-to-GPU communication, it is presented as a fallback for people who see no relief from disabling the service.

My reading is that the service route is the safer first move for most affected machines. FreeSync matters to gamers who rely on variable refresh rates, and hot-plug display issues are annoying, but both are recoverable by re-enabling the service once AMD ships its patch. Turning off Resizable BAR changes a performance feature that many games are tuned around, and the report itself frames it as a last resort.

The diagnostic angle is worth flagging for anyone unsure whether they are affected. The reported symptoms are subtle: no crash, no obvious slowdown, just steady background writes. Tools that report total bytes written per drive, or vendor utilities that surface a health percentage, would show a suspicious upward drift on an idle system. A write rate of 1.5 MB/s works out to roughly 130 GB a day if sustained, which is meaningful for a drive rated in the hundreds of terabytes written. Users who have run a Radeon card since around the middle of the year, particularly recent models like the 9060 XT, are the natural group to check first, though the reports also tie the behavior to the Adrenalin application itself.

What remains unsettled is timing. The acknowledgement is informal, delivered in a Reddit comment with language that promises a fix in a coming release but attaches no date or version. Until that lands, the choice is between living with extra writes and accepting the side effects of the workarounds, and users who disable the service should remember to restore it after updating their drivers. There is also no stated plan from AMD for assessing drives already affected, so anyone with a health reading that has dropped sharply may want to keep records.

Why it matters: Developers and product teams running AMD hardware, including workstations used for builds, rendering or local AI workloads, are exposed to silent, continuous disk writes that shorten drive life without any visible error. Because the trigger is a background service installed with ordinary driver updates, this is a reminder that GPU software bundles carry privileged components that can affect the whole system, and that fleet health monitoring for disk wear is worth having. The practical near-term action is to check drive write counters and health percentages on AMD machines, then decide whether disabling the EEU is an acceptable trade-off until AMD ships its fix; that assessment is my inference from the reported behavior, not a stated AMD recommendation beyond the update comment.

Read at ZDNet · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

AMD has admitted a bug within its External Events Utility is wearing out SSDs, and promises a fix.