ssd swap - high usage of Terabytes Written

[ambient_light]

Uh the formula I used was based on 2%, 60TB, 3mo. Last time I checked 50 * 2% is 100% and 3 mo. x 50 is 150 months with is indeed 12.5 years.

Also where is this 240'000 GB number coming from? That is 240 TB/year which if we go by the 2,400 TBW warranted for a 2 TB Samsung.860 PRO should last 10 years and that is warranted for 5 years.

Right, the historical "% used" extrapolation gives you this, assuming linear wear and usage. What I meant however is the result of the original formula from the article you referred to, FWIW ...where 240TB is the estimated annual TBW.

Also, if we use your Samsung TBW info as a reference, then for 1Tb SSD it gives 2.5-5 years endurance depending on the model (600-1200 TBW) ... not exactly reassuring. Long story short - we don't know the real SSD longevity, calculations above give rather wide range.

And it's not particularly what is interesting here - SSD chips Apple use in M1s aren't much different from what industry is using. The topic is the obvious write abuse by MacOS Big Sur, primarily via swapping, and how to solve it.

 

[IceStormNG]

SSD chips Apple use in M1s aren't much different from what industry is using. The topic is the obvious write abuse by MacOS Big Sur, primarily via swapping, and how to solve it.

Well.. The NANDs are not so much different. They're either WD/Sandisk for the Air and Kyocera (Toshiba) for the Pro. At least that's what iFixit found in their teardown. Maybe there are even more different Chips out there.
The controller however is from Apple and baked into the M1 (or the T2 on current Intel Machines). Controller logic also plays into the equation when it comes to lifetime and wear endurance.

If you look at current NVMe SSDs from WD and Toshiba they're about 600TBW for 1TB (WD) and 400TB for 1TB (Toshiba). This is for their "Pro" models which usually use some better quality NAND. We don't know which Apple used in the M1. At least I couldn't find anything about that online with the chip numbers or from other sources.

 

[jdb8167]

Well.. The NANDs are not so much different. They're either WD/Sandisk for the Air and Kyocera (Toshiba) for the Pro. At least that's what iFixit found in their teardown. Maybe there are even more different Chips out there.
The controller however is from Apple and baked into the M1 (or the T2 on current Intel Machines). Controller logic also plays into the equation when it comes to lifetime and wear endurance.

If you look at current NVMe SSDs from WD and Toshiba they're about 600TBW for 1TB (WD) and 400TB for 1TB (Toshiba). This is for their "Pro" models which usually use some better quality NAND. We don't know which Apple used in the M1. At least I couldn't find anything about that online with the chip numbers or from other sources.

Also, Apple's warranties are structured differently than for someone selling an M.2 NVMe. Apple warranties the whole computer for 1-3 years. They can take a risk and be less conservative about expected TBW. There are numerous reports that in real-world testing, that the stated TBW is much less than the practical lifespan.

 

[nquinn]

Well, where do I start...

1. That is not how things work - non-linearity was explained earlier in this thread (you can do some research)
2. Even assuming that the calculation above gives estimation that is close to reality, not everyone got 2TB, and the actual warranted TBW is not known. Case was reported when M1 SSD died after ca 600 Tb.
3. Finally, even if the SDD eventually survives for few more years, excessive writes of this order of magnitude are not OK. I believe the whole thread is about it, and explanation why shouldn't be required by now...If you're happy with it - fine, there are many people however, who are not

I agree with you. I think people are dismissing this problem too easily. It's especially scary on the smaller capacity models like 256gb and 512gb.

 

[Maximara]

Also, Apple's warranties are structured differently than for someone selling an M.2 NVMe. Apple warranties the whole computer for 1-3 years. They can take a risk and be less conservative about expected TBW. There are numerous reports that in real-world testing, that the stated TBW is much less than the practical lifespan.

I'm curious given the 2014 test which seemed to show the exact opposite. Where can I read these numerous real world reports?

 

[jdb8167]

I'm curious given the 2014 test which seemed to show the exact opposite. Where can I read these numerous real world reports?

What? I've read multiple accounts where testing showed petabytes of writes on drives rated with a TBW in the 100s of terabytes. Where have you seen a rated TBW that was more than the actual lifespan?

 

[Maximara]

What? I've read multiple accounts where testing showed petabytes of writes on drives rated with a TBW in the 100s of terabytes. Where have you seen a rated TBW that was more than the actual lifespan?

Ah I was tired and misread what was posted. I read "stated" as the the results of the 2014 which at the high end resulted in petabytes of being written before the drive went bye bye. This is why I use warranted because TBWs much higher were stated by the 2014 article (which I have cited enough times).

Similarly we have a statement about a drive that failed at only 600 TB which is way below many warranted TBW on the high end

TL; DR: "Stated" and "warranted" do not technically mean the same thing.

 

[jdb8167]

Ah I was tired and misread what was posted. I read "stated" as the the results of the 2014 which at the high end resulted in petabytes of being written before the drive went bye bye. This is why I use warranted because TBWs much higher were stated by the 2014 article (which I have cited enough times).

Similarly we have a statement about a drive that failed at only 600 TB which is way below many warranted TBW on the high end

TL; DR: "Stated" and "warranted" do not technically mean the same thing.

I was confused since you were one of the posters who pointed out the studies! Sorry for the unclear language.

 

[Maximara]

I was confused since you were one of the posters who pointed out the studies! Sorry for the unclear language.

No problem. I tend to read posts like I write them - from a technical prospective. And there the wording has to be clear.

 

[wirtandi]

Those who are on the latest big sur beta, any good news? 11.3.6 or 11.3.7 IIRC....?

 

[tab0reqq]

That it not even remotely true.
View attachment 1758325

You got it wrong. I'm not saying reads and writes are similar. I'm saying that pace of reads and writes in time is similar if you don't change your workflow.

But that's not the case. The case is that if your workflow is very light on RAM, then this topic won't scare you.
But if you try to use the Macbook to its full potential, it is starting to get scary.

 

[TheSynchronizer]

I agree with you. I think people are dismissing this problem too easily. It's especially scary on the smaller capacity models like 256gb and 512gb.

A lot of people in this thread, me included, have also fixed the problem completely. I’ve went from ~250GB written per day to ~35GB writes per day on my M1, and my use has not changed at all. In fact I feel more safe to push the system even harder now!

And I have a 256GB/8GB M1 MBP. But at the rate i’m going my SSD will last ~50 years

 

[rob984]

A lot of people in this thread, me included, have also fixed the problem completely. I’ve went from ~250GB written per day to ~35GB writes per day on my M1, and my use has not changed at all. In fact I feel more safe to push the system even harder now!

And I have a 256GB/8GB M1 MBP. But at the rate i’m going my SSD will last ~50 years

How did you fix it? Browser tweaks or something else?

 

[Thistle41]

How did you fix it? Browser tweaks or something else?

I was thinking that as new thread followers join it is nigh well impossible to extract the fixes. I left a request for a Wiki addition to this thread where the basics of use experienced and how to fix can be posted at the top of the thread.

 

[Grey Area]

A lot of people in this thread, me included, have also fixed the problem completely. I’ve went from ~250GB written per day to ~35GB writes per day on my M1, and my use has not changed at all. In fact I feel more safe to push the system even harder now!

And I have a 256GB/8GB M1 MBP. But at the rate i’m going my SSD will last ~50 years

Great, regarding the lifespan that seems to have solved the problem. 👍

Depending on the usage that still is a remarkable amount of writing going on, though. I mean, I get 35GB a day on our Linux home server that is on 24/7, serving multiple users and both downloading and streaming/buffering stuff constantly. If my personal laptop was writing 35GB a day, I would really want to know what is going on in there.

 

[Maximara]

MBA M1 16/256 (7+8)

It's hard for me to target exact date but I just noticed:

2021.03.13

=== START OF SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED

SMART/Health Information (NVMe Log 0x02)
Critical Warning:                   0x00
Temperature:                        38 Celsius
Available Spare:                    100%
Available Spare Threshold:          99%
Percentage Used:                    1%
Data Units Read:                    35 313 613 [18,0 TB]
Data Units Written:                 33 143 791 [16,9 TB]
Host Read Commands:                 209 435 264
Host Write Commands:                151 145 455
Controller Busy Time:               0
Power Cycles:                       94
Power On Hours:                     136
Unsafe Shutdowns:                   4
Media and Data Integrity Errors:    0
Error Information Log Entries:      0

2021.04.15

SMART/Health Information (NVMe Log 0x02)
Critical Warning:                   0x00
Temperature:                        28 Celsius
Available Spare:                    100%
Available Spare Threshold:          99%
Percentage Used:                    2%
Data Units Read:                    66 653 962 [34,1 TB]
Data Units Written:                 62 347 927 [31,9 TB]
Host Read Commands:                 367 889 424
Host Write Commands:                269 135 121
Controller Busy Time:               0
Power Cycles:                       110
Power On Hours:                     237
Unsafe Shutdowns:                   4
Media and Data Integrity Errors:    0
Error Information Log Entries:      0

So ~
18 TBW -- percentage used 1%
34 TBW -- percentage used 2%

~ 17 TBW = ~ 1%
1700 TBW = ~ 100%

Purchased:
2021-02-05, so 2 months and 2%.

100% = 100 months = over 8 years

That's far from OK if we still can use MBA 2013 without problems (that are 8 years old).
I kinda hope that Apple will tell what the heck. They already lost few battles.

Now this is how one should study this issue.

I should mentioned the percentages given by the program actually range between (x)+0.5 to (x+1)+0.4999 (they are rounded to the nearest integer

18 TBW -- percentage used 1% gives 3600 to 1201
34 TBW -- percentage used 2% gives 2267 to 1360

The second calculation is within the range of the first so this supports the idea of a linear calculation which is backed up by data from The SSD Endurance Experiment: They’re all information. Only one drive had a weird little hiccup and it resulted in the drive lasting longer than the up to than the up to then linear graph would have predicted.

 

[haralds]

If that's the case, then my Windows computer's SSD must have way more TBW down than it actually has.
That machine has 40TBW down in 4 years of heavy usage (generating lots of high res textures assets and video recording).

This is the Preview. I seem to have 1-2 updates a week!

 

[haralds]

Just compared my Mac Pro 5,1 running Big Sur on Open Core. The system disk is a PCiE M2 that has been running about 10 months. Its TBW is now slightly less than my M1 at 5 months. With 70TB the M1 has averaged about 2x the writes of the Intel Mac Pro running the same OS version.
Both machines have a 2TB system drive. The Mac Pro has additional drives for media, repositories, games, and booting other OSs. It also has 48GB RAM vs 16TB on M1.

 

[leons]

I was thinking that as new thread followers join it is nigh well impossible to extract the fixes. I left a request for a Wiki addition to this thread where the basics of use experienced and how to fix can be posted at the top of the thread.

U got it.

 

[Maximara]

Right, the historical "% used" extrapolation gives you this, assuming linear wear and usage. What I meant however is the result of the original formula from the article you referred to, FWIW ...where 240TB is the estimated annual TBW.

Also, if we use your Samsung TBW info as a reference, then for 1Tb SSD it gives 2.5-5 years endurance depending on the model (600-1200 TBW) ... not exactly reassuring. Long story short - we don't know the real SSD longevity, calculations above give rather wide range.

And it's not particularly what is interesting here - SSD chips Apple use in M1s aren't much different from what industry is using. The topic is the obvious write abuse by MacOS Big Sur, primarily via swapping, and how to solve it.

With 60TB written to the drive one would expect 10% (60/600) for "% used" and at 2% we are seeing nowhere near that. Also The SSD Endurance Experiment: They’re all dead shows results that match up best with a linear calculation. Sure it isn't a perfect linear path, and the HyperX had a weird little hiccup in the 600 to 700 TBW range (but that was way above the 80% used range so we can ignore it as it actually extended the drives life) but close enough the method I have been using makes sense.

TL;DR: real world testing supports a linear calculation using percentage used and TBAW (TB Actually Written).

 

[TheSynchronizer]

How did you fix it? Browser tweaks or something else?

Really the most major 'fix' - and the only one I notice making any sort of difference to my use of this mac - was switching from safari to any other browser. Edge, firefox, vivaldi - these are all great options. Currently I'm using the m1 brave browser, so far it's my favourite after more or less trying them all. The most important part is to get a tab discarding extension, such as 'Auto Tab Discard' to not unnecessarily keep every browser tab open at all times. Realistically you're not constantly switching back and forward between more than 10 tabs - are you?

iOS devices do this by default - i.e. the iOS Safari discards background tabs after you open a certain number of them, and only keeps a limited number actively open. We all know that iOS devices have less RAM than any M1 MBP, but essentially this is probably done to not have excessive swapping on the iOS device flash memory (and swapping definitely does occur on iOS because if you fill the storage of an iOS device completely you will notice how it behaves slowly and stuttery being unable to swap).

I have done additional things other than switching browser, which I've documented in this thread in the past, but I'm not going to bother mentioning them again as I believe just switching browser and using tab discarding is enough by itself to fix the writing issues completely - if someone happens to have switched browsers, and still experiences high writes even when using tab discarding, then I'll happily recap every other step I took to reduce writes.

 

[TheSynchronizer]

“The first post of this thread is a WikiPost and can be edited by anyone with the appropiate permissions. Your edits will be public.”

If this means that we can make the first post here by @Forti a summary of steps and fixes to try to fix the writing issue, i’d be happy to contribute a step by step list of things to try which have helped me and others here!

 

[stigman]

guys, I'm going to unpack my MBP 8/256, but this thread makes me a little worried. I've gone through trauma using apple's faulty designed butterfly keyboard and this was the worst experience ever which I had while typing. I wonder what is users mental condition being aware of this issue, but this macbook may contribute to mild anxiety and some sorf of compulsive disorder like checking swap usage or TBW. This is insane.

 

[leons]

guys, I'm going to unpack my MBP 8/256, but this thread makes me a little worried. I've gone through trauma using apple's faulty designed butterfly keyboard and this was the worst experience ever which I had while typing. I wonder what is users mental condition being aware of this issue, but this macbook may contribute to mild anxiety and some sorf of compulsive disorder.

You will love it and you will be happy! Whether or not you are "affected" depends on your usage habits. If you are, there are adjustments you can make. Either way, the effectiveness and joy of using this wonderful machine will not be diminished IMHO.

 

[TheSynchronizer]

guys, I'm going to unpack my MBP 8/256, but this thread makes me a little worried. I've gone through trauma using apple's faulty designed butterfly keyboard and this was the worst experience ever which I had while typing. I wonder what is users mental condition being aware of this issue, but this macbook may contribute to mild anxiety and some sorf of compulsive disorder like checking swap usage or TBW. This is insane.

Are you happy using a browser other than Safari, such as Brave, Vivaldi, Firefox, Edge, with a tab discarding / memory management extension?

If your answer to this question is yes, you likely do not need to worry as you should have years upon years of SSD life.

And in due time Apple will surely fix Safari & make Big Sur swappiness less aggressive, but until then switching from Safari, and if necessary applying a few other tweaks fixes most if not all issues. This is coming from a 8/256 MBP owner who went from insane 250-350GB daily writes, to currently 40GB a day max (usually stays in the 25-35GB range) with even heavier multi tasking than before and all purpose usage. And that calculates to around 50 years of SSD life, at which point I think its safe to relax and not worry anymore