SSD Maintenance

Question

  How Trim Works -
The OS sends a TRIM command at the point of file deletion
The SSD marks the indicated locations as invalid data
TRIM Features:
► Prevents GC on invalid data
► Increases the free space known to the SSD controller
TRIM Benefits:
► Higher throughput – Faster host write speeds because less time writing for GC
► Improved endurance – Reduced writes to the flash
Lower write amplification – Less data rewritten and more free space is available
The OS tracks what files are present and what logical blocks are holding the files
SSDs do not understand the file structure of an OS; they only track valid data locations reported by the OS
When the OS deletes a file, it marks the file’s space in its logical table as free - It does not tell the drive anything
When the OS writes a new file to the drive, it will eventually write to the previously used spaces in the table
An SSD only knows data is no longer needed when the OS tells it to write to an address that already contains data
All SSDs will have some form of GC – it is not an optional feature
NAND flash cannot directly overwrite a page with data; it has to be first erased
One full block of pages has to be erased, not just one page
GC starts after each page has been written one time
Valid data is consolidated and written into new blocks
Invalid (replaced) data is ignored and gets erased
Wear leveling mainly occurs during GC
Garbage Collection (GC) is the process of relocating existing data, deleting stale data, and creating empty blocks for new data
TRIM SSD and GC
The SATA 3.1 specifications Queued Trim Command - allows SATA SSDs to execute Trim without impacting normal operation...
http://www.sata-io.org/technology/6Gbdetails.asp
Wikipedia SSD:
"includes support for the TRIM command to reduce garbage collection for data which the operating system has already determined is no longer valid. Without support for TRIM, the SSD would be unaware of this data being invalid and would unnecessarily continue to rewrite it during garbage collection causing further wear on the SSD."
https://en.wikipedia.org/wiki/Solid-state_drive
https://en.wikipedia.org/wiki/TRIM
http://en.wikipedia.org/wiki/Write_amplification

Answer 1

Russ Arcuri - Macintouch commentary
RE: Michael Diehr's commentary about TRIM.
I can only assume Michael was not around for the previous discussions about TRIM here, during which each and every objection he raises was hashed out.
So-called "sophisticated" or "advanced" garbage collection only exists because TRIM-like technology didn't exist already, being unnecessary for use with spinning magnetic media. Something had to be done to allow SSDs to work properly with drive interfaces, filesystems and operating systems that were optimized for use TRIM directly addresses the shortcomings of having only garbage collection available. SSD controller manufacturers and designers (including SandForce, the controller manufacturer for OWC's SSDs), recommends that TRIM be used with their products. So does Samsung.
For example, here's a 2011 article from OWC describing how you don't need TRIM on their SSDs and how it can in fact hurt performance or reliability.
That article has been discussed here on MacInTouch before. In my opinion it's bad advice, and inaccurate in some of its assertions. It also ignores the recommendation made by SandForce to use TRIM with their SSD controllers. But even if one were to take that article at face value, applying that advice to SSDs other than OWC's makes little sense.
The reason I'm advising against TRIM is simply that it's yet another driver-level modification of the OS, and these always carry potential risk (as all the folks with WD hard drives who lost data can attest to).
Apples and oranges comparison, for a variety of reasons. The short of it is that TRIM is supported natively in all recent versions of OS X. The tools used to enable it for third party SSDs do not add a new kernel extension; they change the setting to allow Apple's native TRIM implementation to be used with SSDs other than those factory installed by Apple.
This shows that the 840s do work slightly better with TRIM than without, but the differences are (in my opinion) trivial, a 9% increase at best.
One of the major reasons for the skepticism that exists about TRIM is that so many people, the authors of both articles you linked to included, don't understand it.
TRIM is not, strictly speaking, a performance-enhancement technology -- though it is plainly obvious that most people think it is.
Though it can, in many circumstances, improve performance, there are also circumstances under which it will provide little or no noticeable benefit. Not coincidentally, a new SSD tested fresh out of the factory packaging is unlikely to show much (if any) benefit. Or rather, TRIM is providing a real benefit for new SSDs, but that benefit doesn't become measurable in terms of benchmark performance testing until every memory cell in the SSD -- including many gigabytes of cells hidden from visibility by the SSD controller -- have been written to at least once. Writing 128 GB of files to an SSD with a nominal capacity of 128 GB won't do it, as there are several gigabytes (exact number varies depending on the model) still unwritten.
Under real-world use conditions, having TRIM disabled means eventually having noticeable write performance degradation due to write amplification. It is far greater than "9%" -- it can be a 50% or greater drop in write performance, depending on various factors. Defining "eventually" is difficult because it depends on how the SSD is used. But given enough time and write cycles, it can happen to all SSDs used without TRIM, no matter how sophisticated their garbage collection algorithms are.
Under those same real-world use conditions, having TRIM enabled means that the SSD should almost never reach a state of having noticeable write performance degradation, as it should almost never get into a state where write amplification is happening.
I will concede that it is possible to design a lab test in such a way as to defeat the benefits provided by TRIM, but such tests do not reflect any real-world usage scenario I can imagine. Furthermore, those same contrived tests would put an un-TRIMmed drive into an equally-addled state even more quickly.
I would suggest reading through the rather lengthy previous discussions about TRIM. Here are a couple of my past posts that are most relevant to the current discussion:
A description of what TRIM is here.
I addressed some of OWC Larry's comments about TRIM use with OWC/SandForce SSDs here.
http://www.macintouch.com/readerreports/harddrives/index.html#d09dec2013