This looks alarming. I hope Apple issues a a fix soon. I have a M1 MacBook Pro with only 8GB of ram.
Last edited:
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
M1 Mac Users Report Excessive SSD Wear
- Thread starter MacRumors
- Start date
- Sort by reaction score
smartmontools download
Edit: Unfortunately, the standard download is x86 only. You will have to build or use brew to get an Apple Silicon version.
Last edited:
It's so concerning, puts a shadow over my whole M1 experience. I think I will have to stop using my M1 device moving forward.
I dunno. I still use my 11-year-old-bought-on-launch-day-one Macbook air daily.
It's mind boggling that this thread has gotten so much traction. No one is actually taking the time to calculate this disk usage in years and skipping over all the comments that have done the math. Using 1% health every 2 months would equate to over 16 years. Twice that usage would still be 8 years.
Also, the M1 is literally the same tech in the iPhones and iPad Pros and I have seen no reports of people complaining about iPads having massive disk failure. Those devices don't even have as much RAM. They also don't typically share all the same utility data that people monitor on PC's hence people never worry about it. This whole thread seems to be based more on speculation than fact.
Not sure this concern merited a front-page Macrumor article just to propel mass hysteria. Also, as much as people are revering disk platter drives in these comments, they forget how easily they broke/died for tons of other reasons other than read/write cycles.
Also, the M1 is literally the same tech in the iPhones and iPad Pros and I have seen no reports of people complaining about iPads having massive disk failure. Those devices don't even have as much RAM. They also don't typically share all the same utility data that people monitor on PC's hence people never worry about it. This whole thread seems to be based more on speculation than fact.
Not sure this concern merited a front-page Macrumor article just to propel mass hysteria. Also, as much as people are revering disk platter drives in these comments, they forget how easily they broke/died for tons of other reasons other than read/write cycles.
This is FUD.
If the numbers being reported here are true (6% wear in 2 weeks some people are suggesting), then there would be a video in the next 3 days that will be able to stress test the SSD and successfully break it, especially when people are saying that it is not that hard to do 100 TB worth of write + read.
This will not be a theory in 3-5 days as it will either be debunked or be proven.
If the numbers being reported here are true (6% wear in 2 weeks some people are suggesting), then there would be a video in the next 3 days that will be able to stress test the SSD and successfully break it, especially when people are saying that it is not that hard to do 100 TB worth of write + read.
This will not be a theory in 3-5 days as it will either be debunked or be proven.
Except I don't think anyone knows why the writes are so high so it might be difficult to manufacture a failure.
I (MBP M1 16GB RAM, BigSur) compared TBW result with my friend (MBP 2018 32GB RAM, BigSur).
for this "test" I made sure that Im not getting my SWAP used - and while I got ~20GB data written in one day just from other tasks I do, he got just 200MB written in 3 days!
so, based on this I would assume that the RAM is not solely a problem, but it adds its five cents to it.
what was odd for me, is that my TBW "jump" sort of matched what was visible in Activity Monitor data written summary (few gigs from kernel_task, few from launchd and few from my PHPStorm IDE), but in my friends Intel MBP case, his TBW reported by smartctl wasnt equal to what Activity Monitor reported (he had ~80GB in activity monitor, while only 200MB difference in TBW in smartctl between two days).
not sure to what conclusion this leads, but for me it looks like its not RAM that is the main cause, but its something else that either write too much to disk or maybe reports wrong numbers? (but whether its on M1 BigSur or Intel BigSur?)
for this "test" I made sure that Im not getting my SWAP used - and while I got ~20GB data written in one day just from other tasks I do, he got just 200MB written in 3 days!
so, based on this I would assume that the RAM is not solely a problem, but it adds its five cents to it.
what was odd for me, is that my TBW "jump" sort of matched what was visible in Activity Monitor data written summary (few gigs from kernel_task, few from launchd and few from my PHPStorm IDE), but in my friends Intel MBP case, his TBW reported by smartctl wasnt equal to what Activity Monitor reported (he had ~80GB in activity monitor, while only 200MB difference in TBW in smartctl between two days).
not sure to what conclusion this leads, but for me it looks like its not RAM that is the main cause, but its something else that either write too much to disk or maybe reports wrong numbers? (but whether its on M1 BigSur or Intel BigSur?)
uh.. what?
"Software ensures that load is spread evenly across the drive's memory cells"
My understanding is its the controller of the flash storage to deal with this, *in addition*to usual garbage collection not any other external factor.. Apple does that separately in software?
example: an OS has no idea what cell the flash will write to, its all handled by the logic's of the SSD. Very similar to how LBA translation works on spinning hard drives.
"Software ensures that load is spread evenly across the drive's memory cells"
My understanding is its the controller of the flash storage to deal with this, *in addition*to usual garbage collection not any other external factor.. Apple does that separately in software?
example: an OS has no idea what cell the flash will write to, its all handled by the logic's of the SSD. Very similar to how LBA translation works on spinning hard drives.
This thread is no more than an invitation to hate for the “I’ll never buy a first gen Apple product, I love my Intel, I can’t wait for the smallest unproven M1 issue Twitter post so I can brag about my hate for M1 computers” crowd.
End of story. Nothing to see here. I’d bet my net worth 5 years from now, there will be no more M1 SSD failures than Intel SSD failures. In the meantime, I’ll enjoy my speed demon of an M1.
End of story. Nothing to see here. I’d bet my net worth 5 years from now, there will be no more M1 SSD failures than Intel SSD failures. In the meantime, I’ll enjoy my speed demon of an M1.
Guess you mean "buy a new MacMini with more RAM if you need it."?
That's the EXACTLY reason why we need removable SSD and replaceable parts! Apple has shown its tiresomely and irritatingly repetitive committment to the environment, but this really does opposite, forcing us to buy new Macs when SSD or soldered-on parts fail.
I spent €200 on the new GPU card to replace the idiotic AMD Radeon 5750 GPU in my 10-year-old iMac 27" (Mid-2010) so I can install the Big Sur and run apps that use Metal engine. I see no need to buy new iMac when my current iMac runs beautifully to this day.
I spent €200 on the new GPU card to replace the idiotic AMD Radeon 5750 GPU in my 10-year-old iMac 27" (Mid-2010) so I can install the Big Sur and run apps that use Metal engine. I see no need to buy new iMac when my current iMac runs beautifully to this day.
The problem with your close minded statement is the fact that SSD’s aren’t failing. You are complaining about a problem that doesn’t yet exist. Do you complain that your hamburger is burned before you start cooking it as well?
I'm not seeing what others see either on my 2014 13" Retina MBP running iStat Menus version 6.51. Rather than seeing the lifetime left under Status I see "S.M.A.R.T Status Passed" ......
