Comments on: Trimming SSD Performance Degradation https://www.thessdreview.com/ssd-guides/optimization-guides/ssd-performance-loss-and-its-solution/ The Worlds Dedicated SSD Education and Review Resource | Mon, 18 Nov 2013 13:02:00 +0000 hourly 1 By: Alan https://www.thessdreview.com/ssd-guides/optimization-guides/ssd-performance-loss-and-its-solution/#comment-16670 Mon, 18 Nov 2013 13:02:00 +0000 https://thessdreview.com/?p=3253#comment-16670 Hi Les,
I understand the concept and value of overprovisioning but not how best to implement it. Some writers say it is enough to just leave empty space on the drive in the same partition as the OS whilst others say it is necessary to make an unallocated partition which will then be used by the controller as overprovisioned space. I wonder if you could clarify this. I am using OCZ Vectors.

]]> By: Les@TheSSDReview https://www.thessdreview.com/ssd-guides/optimization-guides/ssd-performance-loss-and-its-solution/#comment-16667 Sun, 17 Nov 2013 19:40:00 +0000 https://thessdreview.com/?p=3253#comment-16667 In reply to jameskatt.

Unfortunately I am not an expert, nor would I jump at that designation for many in this industry but you answer your question right off when you spoke of garbage collection in SF drives.

]]> By: jameskatt https://www.thessdreview.com/ssd-guides/optimization-guides/ssd-performance-loss-and-its-solution/#comment-16666 Sun, 17 Nov 2013 19:27:00 +0000 https://thessdreview.com/?p=3253#comment-16666 In reply to Les@TheSSDReview.

Since SSDs sooner or later are going to take over Rotational Hard Drives for external USB 3.0 general purpose storage, it should be true that TRIM is no longer needed. USB 3.0 drives will not receive TRIM commands even if TRIM is enabled.

Since you test SSDs on this site, I think you would be most expert in determining if SSDs installed in external USB 3.0 enclosures have performance degradation over time since they won’t be capable of receiving the TRIM command.

]]> By: Les@TheSSDReview https://www.thessdreview.com/ssd-guides/optimization-guides/ssd-performance-loss-and-its-solution/#comment-16665 Sun, 17 Nov 2013 19:21:00 +0000 https://thessdreview.com/?p=3253#comment-16665 In reply to jameskatt.

Although a dated article, the argument is a valid one, however, is primarily based on LSI SandForce controlled SSDs in this argument. Can your same opinions be reflected for all manufacture of SSDs today?

]]> By: jameskatt https://www.thessdreview.com/ssd-guides/optimization-guides/ssd-performance-loss-and-its-solution/#comment-16664 Sun, 17 Nov 2013 19:05:00 +0000 https://thessdreview.com/?p=3253#comment-16664 The biggest argument I have against enabling TRIM is that TRIM is a SATA-only command.

SSDs attached via PCIe, USB 3.0, FireWire, or Thunderbolt cannot receive the TRIM command. Macs won’t even recognize them as SSDs nor would the Mac send them the TRIM command even if it is enabled.

The newest MacBooks now use PCIe SSDs.

SSDs (such as OWC Mercury SSDs with SandForce controllers) have evolved to the point they don’t need TRIM – particularly when the SSDs are not attached via SATA? They do their own garbage-collection and optimization.

OWC – in particular – advises against enabling TRIM on their Sandforce controller OWC Mercury SSDs since this increases wear and tear on their SSDs. TRIM adds extra unnecessary writes when the SSD already did this on its own. Enabling TRIM would essentially harm the SSD and shorten its lifespan. https://blog.macsales.com/11051-to-trim-or-not-to-trim-owc-has-the-answer

If modern SSDs need TRIM, they would be in danger of serious performance problems if the SSD was attached via a non-SATA connection such as USB 3.0 or PCIe or Firewire or Thunderbolt. Unless the manufacturer insures their SSD’s controller does its own version of TRIM, then they would suffer serious performance problems when attached via non-SATA connection compared to the competition.

This is why I argue that TRIM is currently useless in modern SSDs.

]]>