Assumption about SSD & GC & TRIM

[hhhd1]

Is it true that if you run an SSD in a system with no TRIM, but leave some space for extra over-provisioning, that the SSD would not be affected by speed degradation due to lack of TRIM ?

Because the SSD will always have an extra space to be used when it receives new writes, and the SSD will always know that this space is always unallocated ?

Crucial adrenaline is just a Crucial M4 with 20% over-provisioning, correct ?

 

[ShintaiDK]

TRIM got zero effect on degrade. TRIM only optimizes the write speed slightly.

So you get the same lifespan out of an SSD with or without TRIM.

 

[hhhd1]

Yes, I am talking about write speed degradation only, not life span.

(I think that TRIM does help with life span, because it gives the ssd more free room to move stuff more efficiently, therefore decreasing WA, but that is a story for a different thread)

 

[ShintaiDK]

You can see the article on Anandtech about TRIM vs no TRIM if I recall right. Its not exactly something that raises concerns.

I think best case gave TRIM something like 25% advantage. But only if you had just witten a huge amount, deleting it and written it again. Everyday usage is more like 5% I bet.

TRIMs main benefit is the time before GC kicks in vs TRIM.

 

[Cerb]

Yes, I am talking about write speed degradation only, not life span.

(I think that TRIM does help with life span, because it gives the ssd more free room to move stuff more efficiently, therefore decreasing WA, but that is a story for a different thread)

If TRIM's impact on WA is more than negligible, nobody should be buying that SSD, because that shows they did a very poor job of designing and testing it (poor WA algorithms, not enough reserved space, etc.). Luckily, that has been a largely moot issue for awhile now, unless you happen upon some old bad controller.