Two systems and lifetime increase

Hello! Need your help!
3 days ago I bought iMac late 2013 with new generation PCIe Samsung XP941 M.2 based SSD 256Gb.
Really powerful SSD. But of course, I was scared about SSD rewrite cycles and wanted to check it because want to keep this computer for 5 years. Found Mac OS X app called DriveDx that can check SSD and HDD lifetime and other S.M.A.R.T. attributes . Utility log here:

SSD Lifetime Left Indicator : GOOD 95.0%
=== SSD HEALTH INDICATORS ===
ID NAME RAW VALUE STATUS
5 Retired Block Count 0 100% OK
173 Wear Leveling Count 0x100990005 95.0% OK
174 Unexpected Power Loss Count 0xA368E 99.0% OK
192 Unsafe Shutdown Count 6 99.0% OK
197 Current Pending Sectors Count 0 100% OK
199 UDMA CRC Error Count 0 100% OK

I used this Mac for 3 days. I just did clean install OS X , installed 80 GB of soft on it, and installed Windows in the second partition and 95% of wear leveling count?!
My SSD has 2 partition: Mac OS X (133 GB) and second for Windows 8.1 (117 GB)
I know that OS X is very optimized to SSD, but I have second Windows partition and don’t know how two systems will live together. Need Windows ONLY for gaming. I have two question.

First:
I heard that rewrite cycles depends on SSD size. Is it true that my drive turned into two 128 Gb SSD? When one of OS reach 128 Gb then it will rewrite cells again. Right? And lifetime will be a half of 256 Gb. And how about TRIM with two different systems. Is it working correctly?

Second:
I have an idea how to reduce read/write on Windows. Want to make 30Gb partition for Windows and install all games in my external HDD USB 3.0. Yeah, game loading will be slower, but less read/write gaming data. But this method is useless if external HDD use system SSD as temp storage between RAM and HDD.
Does Windows use system drive to store some temporary data between system drive and external HDD or it has direct route to RAM from HDD? If it does, using external HDD is useless, because all temp data will be passing through system SSD and make rewrite.

 

You're overthinking this.

Your setup has two partitions on your single SSD and the SSD controller puts files where it needs to (to optimize the longevity of the nand cells).


Don't know about what that software is reporting (all it can do is guess, after all) but if you want to give your SSD the best chance of giving you the longest lifetime possible and the highest performance at the same time, do this:

Shrink one of the partitions so that you leave 30% or so of the total drive capacity as 'unallocated'. Doesn't matter where you leave this space as 'unallocated' - the SSD's controller will make use of this capacity to ensure the least WA (write amplification) during TRIM and GC (garbage collection) routines. And the side benefit will be a snappier experience in either O/S you use.


Hope this helps.

Good luck.

 

its more likely that a human error occurs that causes the SSD to completely fail without any signs of wear than an SSD actually going through a lot of data writes per day for the following 5 years to deplete its write cycles

3000 writes on a single cell is not a lot but 3000 times 256Gb is about 800 terabytes (768), I'll do the calculations for you

say you want the SSD to die in 5 years, you'll need to write: 160 terabytes a year, or 13.3 terabytes a month, or 400Gb a day, and IF i'm being wrong for the 3000 write cycles, change it to 1000 and its still 133Gb a day, yes I can READ 133Gb data a day but I'd barely touch 30Gb writes a day even with torrent software running occasionally for anime downloads

conclusion: you break the SSD yourself during a power surge before it writes over its life cycle

 

Just because you can write to it, doesn't mean it can hold the data and allow you to read it back as it was intended (the 'specs' for an SSD is to be able to read data reliably after a year of no power applied).

Not only that; but it is not just the fact of writing to an SSD that determines it's lifespan (simply reading an SSD wears it out too). You have to consider the type of write (or workload) it is subject to. Sequential writes may give you those 800TB numbers people like to quote - but random 4K writes will kill a consumer SSD in a matter of weeks.


Point: multiplication doesn't give us an estimate of an SSD's lifespan - as long as the SSD is used as an O/S and program drive, the 'type' of workload it has is far from sequential, overall.