Amazingly terrific if reported correctly by the system. Congratulations.
Amazingly terrific if reported correctly by the system. Congratulations.
My 500GB SSD is showing 15.1TB written, and is still on 0% used.If you are at all comfortable with the terminal command line you don't need to pay anything since smartmontools is open source and free. You can install it with Homebrew (https://homebrew.sh).
Install
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
Then
brew install smartmontools
I'm at almost 8 TBW now and still at 0% used (I don't think that the % used has been verified to be correct yet but it is the best we have right now). So assuming it goes to 1% at 8 TBW (no evidence of this but just using it as a guess) then 100% would be 800 TBW. This is on a 1 TB SSD. So 1 TBW/month would be 800 months or 66 years. I think you probably don't have to be that conservative. Note that smaller SSDs tend to have shorter TBW values but even 400 TBW would be 33 years. The people who are having problems are seeing 1 TBW in a matter of days not months.
OK. I'm 900GB away from that number on my 500GB SSD so I'll see what happens. What SSD size are you running this test on?I've got good news for both of you (and all)........1% is 16TBW.......NOW recompute!
OK. I'm 900GB away from that number on my 500GB SSD so I'll see what happens. What SSD size are you running this test on?
[Update: just saw that @leons is using a 256GB SSD. So maybe I would see 1% at 32TBW?.
My MBP16 with 1TB SSD shows 2% at 172TBW....which as @Maximara pointed out would result is an impossibly large TBW for a consumer SSD (over 8000TBW)...so we need to take the "% used" figure with a large pinch of salt.
A little digging on the old internets and I found a twitter post by Dan Moren showing smartctl output at 17.7 TBW showing 1% on a 256 GB SSD. So at least by those numbers from someone known to be knowledgeable and reliable, it appears you are correct. Thanks.
Yes, & BTW, my numbers and percentages are not coming from ANY tool. They are system SMART values being diretly extracted. I don't believe that Apple falsified (or allowed such) the SMART values on the drive, or there would be a class-action suit of epic proportions and we will all be rich!A little digging on the old internets and I found a twitter post by Dan Moren showing smartctl output at 17.7 TBW showing 1% on a 256 GB SSD. So at least by those numbers from someone known to be knowledgeable and reliable, it appears you are correct. Thanks.
https://twitter.com/dmoren/
Congratulations to ALL of us!Amazingly terrific if reported correctly by the system. Congratulations.
You are correct, 32TBW. Make a calendar entry to let us know!OK. I'm 900GB away from that number on my 500GB SSD so I'll see what happens. What SSD size are you running this test on?
[Update: just saw that @leons is using a 256GB SSD. So maybe I would see 1% at 32TBW?]
Change to Edge or Firefox and your writes will be a fraction of that.My 500GB SSD is showing 15.1TB written, and is still on 0% used.
In my experience a lot of the writes are simply a function of having significant swap usage. The effect appears to be more noticeable with over about 5GB swap in use. Safari seems to be the biggest culprit, and I am interested in the possibility that changing browser (and disabling cache) could be a way of reducing memory usage.
On a busy day (when I have a lot of necessary tabs open) I can see up to 500GB written per day. If I take a conservative approach and shut tabs and apps down when not using them, then it can be as low as 50GB per day.
The smartmontools are also extracted from SMART values from the drive controller. I've looked at the smartmontools code since it is open source. It is pretty easy to see the values come from an Apple library. The extracted block of data is not documented by Apple but the NVMe spec is available and matches the values in the byte array. So if the % used value isn't correct, that is on Apple for either not estimating correctly or for not following the NVMe specification. I've verified to my satisfaction that the actual TBW values are being correctly reported at least for my use.Yes, & BTW, my numbers and percentages are not coming from ANY tool. They are system SMART values being diretly extracted. I don't believe that Apple falsified (or allowed such) the SMART values on the drive, or there would be a class-action suit of epic proportions and we will all be rich!
The fact that Apple has been silent on this is actually comforting. If the numbers WERE incorrect, with the many people and postings complaining about this, Apple's lawyers would never let them ignore the issue, as it would increase their liability and culpability in a future lawsuit.The smartmontools are also extracted from SMART values from the drive controller. I've looked at the smartmontools code since it is open source. It is pretty easy to see the values come from an Apple library. The extracted block of data is not documented by Apple but the NVMe spec is available and matches the values in the byte array. So if the % used value isn't correct, that is on Apple for either not estimating correctly or for not following the NVMe specification. I've verified to my satisfaction that the actual TBW values are being correctly reported at least for my use.
How many tabs do you typically run?Safari has gotten better at memory management and caching from my use in 11.3 beta 5. Will report if anything changes in beta 6. I hope the behaviour is the same or gets better, as I truly love Safari.
Around 10-15 when I'm doing research and additional tabs for youtube.How many tabs do you typically run?
Sounds good. Those of us in the 20-50+ Tab Camp will have to wait for one of our kind to bravely try out beta 5.Around 10-15 when I'm doing research and additional tabs for youtube.
My SWAP usage is typically < 200 MB when using Safari on my workload. I see around 20 GB written daily on top of everything else I do which is perfectly fine with me. Granted, I am only a student and not doing anything too intensive compared to most users here!
Yeah, I imagine that most will also have some sort of pro apps they may use on top of Safari (or whatever browser) as well. I only use Word and FCP when I need to do edits on videos (usually 4k).Sounds good. Those of us in the 20-50+ Tab Camp will have to wait for one of our kind to bravely try out beta 5.
Most of the time only 1. There are times where I have 2 tabs going, but not too often.How many tabs on average?
There is no "should". At the beginning of your day, run Activity Monitor. Select the Disk Tab, and write down the "Data Written" number on the bottom right. Don't reboot your computer between readings. 24 hours later, do the same, do the math, and report back (assuming it is a "typical day", otherwise, do it again). Let us knowHello. This is my first post here. I've got a 16GB Macbook M1 three days ago and I've been reading this thread carefully. A huge thanks to the whole community (and especially to leons and TheSynchronizer) for the useful advices!
There's one thing which isn't clear for me - what is a "good" amount of swapped memory? And how many GB "should" I write on a daily basis?
I noticed when running some games on Parallels (while keeping Firefox open on the Mac side) I'm getting 1-2GB of swapped memory. Is this "too much"?
Swap itself, or the size of it, aren’t in themselves the issue here. A high frequency of swap in/outs are the main contributors to excessive SSD writes, so they are somewhat linked but regardless not that useful.Hello. This is my first post here. I've got a 16GB Macbook M1 three days ago and I've been reading this thread carefully. A huge thanks to the whole community (and especially to leons and TheSynchronizer) for the useful advices!
There's one thing which isn't clear for me - what is a "good" amount of swapped memory? And how many GB "should" I write on a daily basis?
I noticed when running some games on Parallels (while keeping Firefox open on the Mac side) I'm getting 1-2GB of swapped memory. Is this "too much"?
That gives 1600 TBW for 100% for the medium. The range is a ridiculous 3265 (0.5%) on the high end and 1067 (1.4999...) on the low end. That is an insane margin of error.I've got good news for both of you (and all)........1% is 16TBW.......NOW recompute!
Even given the margin of error, I'll take the 100% on the low side! and not complain!That gives 1600 TBW for 100% for the medium. The range is a ridiculous 3265 (0.5%) on the high end and 1067 (1.4999...) on the low end. That is an insane margin of error.
I've also disabled by OneDrive client sync app. The main reason is because it simply doesn't appear to handle syncing large sets of files (I have over 400,000), even with "offline files" enabled. It tries to process my 400,000+ files at the rate of about 1 per 5 seconds (> 23 days). It also seems to do a lot of disk writing (500GB per day?) that I can do without....
I have no idea what OneDrive is trying to process, when it should not be downloading or uploading any files with "offline files" enabled. I suppose it's just scanning every file to crate meta-data for its sync process, but this might involve actually downloading the file temporarily to scan it? Who knows, but it's a incredibly frustrating piece of software on both Mac and Windows, but especially on Mac.
I have 1TB storage with my MS business account, but I'm going to use it for long term archive only rather than real-time syncing. Dropbox and Google Backup & Sync are *way* faster for the actual sync. OneDrive is quite fast if you up/download directly to the browser or use a 3rd party file transfer tool like Cyberduck.
I was wondering whether to do that as well!I've also disabled by OneDrive client sync app. The main reason is because it simply doesn't appear to handle syncing large sets of files (I have over 400,000), even with "offline files" enabled. It tries to process my 400,000+ files at the rate of about 1 per 5 seconds (> 23 days). It also seems to do a lot of disk writing (500GB per day?) that I can do without....
I have no idea what OneDrive is trying to process, when it should not be downloading or uploading any files with "offline files" enabled. I suppose it's just scanning every file to crate meta-data for its sync process, but this might involve actually downloading the file temporarily to scan it? Who knows, but it's a incredibly frustrating piece of software on both Mac and Windows, but especially on Mac.
I have 1TB storage with my MS business account, but I'm going to use it for long term archive only rather than real-time syncing. Dropbox and Google Backup & Sync are *way* faster for the actual sync. OneDrive is quite fast if you up/download directly to the browser or use a 3rd party file transfer tool like Cyberduck.