Micron 5200 and 5100 SSD slowdown at 65,535 power-on hours: affected MTFDDAK part numbers, Dell and Cisco firmware fixes, and when the 5300 and 5400 are at risk

Micron 5100 and 5200 SATA SSDs that pass 65,535 power-on hours can slow to a crawl while SMART still reports them healthy. Here are the affected Micron and OEM part numbers, the fixed firmware Dell and Cisco now ship, how to check a fleet, and why the 5300 and 5400 have dates of their own.

A storage fault that hides from every health check has been doing the rounds on practitioner forums, community blogs and vendor support sites over the last few weeks. Operators running older servers report SATA SSDs that suddenly deliver a fraction of their normal throughput, pass SMART, show nothing in the controller logs, and come back to full speed for about an hour after a power cycle. The common factor is the Micron 5000 series SATA SSD, a common boot and data drive in 13th and 14th generation Dell PowerEdge, Cisco UCS, HPE ProLiant and white-box servers. Dell updated its knowledge base article on 2 October and Cisco published a field notice on 21 September, so this is now a documented, fixable defect rather than a forum rumour.

What happens at 65,535 hours

• The drive firmware keeps a workload log entry counter in a 16-bit field. It reaches its ceiling of 65,535 at the same point as 65,535 power-on hours, which is just under seven and a half years of continuous running.
• Once the counter is full, the firmware writes debug messages on every update cycle. That logging uses controller CPU time that should be serving I/O, so reads and writes slow down.
• Cisco puts the drop at between 50 percent and more than 90 percent depending on workload. One operator who published a write-up reported throughput falling to around 1 MB/s across a 16-drive array.
• Dell and Cisco both state there is no data loss and no drive failure. The drive stays online and SMART stays clean, which is exactly why it gets misdiagnosed as a controller, cable or hypervisor fault.
• Power cycling the drive or server restores full speed temporarily. Cisco says the slowdown can return roughly an hour later, so it is a diagnostic test, not a fix.
• Updating to fixed firmware cures it, including on drives that have already crossed the threshold.

Which Micron drives are in scope

Cisco's defect notes list all four generations as affected and give the April 2027 and November 2029 dates for the 5300 and 5400. Dell says it is developing firmware for the 5300 and 5400. Fixed firmware is shipping for the 5100 and 5200 now.

Dell PowerEdge: fixed firmware by Dell part number

• Dell's affected-products list runs from the R630, R730 and R730xd through the R640, R740, R740xd, R6515, R7525 and R750 to current R760, R7625 and R770 systems, because the same drives were offered as options across several generations.
• Dell publishes the firmware packages through its normal support downloads for each server.

Cisco UCS and HyperFlex: field notice FN74426

• Affected Cisco product IDs include UCS-M2-240GB, HX-M2-240GB, UCS-SD240GBMS4-EV, UCS-SD480GBMS4-EV, UCS-SD19TBMS4-EV, UCS-SD76TBMS4-EV, UCS-SD19TM1X-EV, UCS-SD38TM1X-EV and HX-SD38TBE1NK9.
• Fixed drive firmware is D0MH079 for most of the list, D0MH449 for UCS-SD38TM1X-EV, D0MH849 for UCS-SD76TBMS4-EV and D0MC448 for HX-SD38TBE1NK9.
• First fixed server bundles: UCS C-Series standalone 4.2(3r), UCS Manager B-Series and C-Series 4.2(3s), Intersight UCS X-Series 5.0(4l).
• Cisco warns that on HyperFlex the extra latency can disrupt Zookeeper transactions, which can take a cluster down. HyperFlex owners should treat this as urgent rather than a performance nuisance.

HPE, Lenovo, Supermicro and Fujitsu systems

• OEM-branded drives take OEM firmware. Use the server vendor's drive firmware package rather than Micron's own image, and check the vendor's support site for an advisory against your drive's model string.
• Drives running Micron's own firmware, which is common in Supermicro and Fujitsu builds, are the awkward case. Operators report that the 5100 and 5200 are past Micron's support window and the fixed image is not on Micron's public download page, which still lists firmware for the 5300.
• One operator who wrote up the fix could not flash the drives through the hardware RAID controller and had to update them one at a time on a direct SATA connection. Budget downtime for that if your drives sit behind a MegaRAID or Smart Array controller.
• This is not the first hour-counter bug in enterprise SSDs. HPE has previously issued a critical bulletin for certain SAS SSD models that failed at 40,000 hours of operation, so it is worth putting power-on hours on every drive audit, whatever the brand.

How to check a fleet in an afternoon

• Pull the model string and firmware revision for every SATA SSD from iDRAC, iLO, XClarity, Intersight or smartctl. Any model starting MTFDDAK or MTFDDAV with a TBY, TCB, TCC, TDC, TDD or TDN code is a 5100 or 5200.
• Read power-on hours (SMART attribute 9). Anything past about 60,000 hours is either already affected or will be within months.
• Compare the firmware against the fixed versions above. A drive on fixed firmware is safe regardless of its hours.
• If a server has unexplained storage latency and healthy SMART, power cycle it as a test. Full speed for about an hour, then a slowdown, points at this bug.
• Log the 5300 drives now. The first of them can reach the threshold from April 2027, and Dell does not yet have fixed firmware for them.

The ETON view

This bug lands squarely on the second-user market. The 5100 and 5200 were default SATA options in the R630, R730, R640 and R740 generation and in a lot of UCS and ProLiant builds from the same years, and those servers are now the backbone of hosting estates, labs and refurbished resale. A drive with 66,000 hours can test perfectly and still be the reason a VM host feels slow. If you are buying refurbished servers or drives, ask for the power-on hours and firmware revision on every SATA SSD before you accept them, and expect the price to reflect a drive that is past seven years of service.

For existing estates the cheapest fix is firmware where the vendor provides it. Where it does not, typically Micron-firmware drives in Supermicro or Fujitsu servers, replacement is usually quicker than a manual reflash across dozens of hosts. Swapping a 5200 for an older-stock 5300 only moves the deadline to 2027, so we would point mission-critical replacements at current drives on the server vendor's own qualified list and keep the 5300 and 5400 on a power-on hours watch.

We source OEM-labelled Dell, HPE, Cisco and Lenovo drives as well as generic Micron parts, so we can quote a like-for-like OEM swap next to a generic option for the same bay and tell you which needs fixed firmware. When the old drives come out, we buy them back.

Related infrastructure

• Hard Drive Options
• Dell PowerEdge Rack Servers
• Dell VxRail HCI
• Cisco M.2 and boot devices
• Disk Storage & Enclosures

Category: Storage · Vendor: Micron, Dell, Cisco · Technology: SATA SSD, Micron 5100, Micron 5200, Micron 5300, Micron 5400, SSD firmware · Last verified: 09 Oct 2026 · ~7 min read

Sourcing this kind of infrastructure? Talk to ETON about availability, lead time and pricing.