SSD - AHCI vs IDE

[Arg Clin]

I just made the switch from IDE to AHCI on my Corsair F120 / AMD 890FX. Performance gains seems impressive, to the point where I'm beginning to doubt that I didn't mess up somewhere:

Crystaldiskmark 3.0.1 x64 IDE:

Seq: Rd 156,8 Wr 82,93
512K: Rd 153,4 Wr 80,20
4K: Rd 17,53 Wr 64,68
4K QD32: Rd 20,82 Wr 84,75

Crystaldiskmark 3.0.1 x64 AHCI:

Seq: Rd 200,8 Wr 86,17
512K: Rd 189,4 Wr 82,07
4K: Rd 17,09 Wr 49,43
4K QD32: Rd 115,0 Wr 78,69

Not increase on every count but some nice gains on the read side. 4K QD32 is close to 6x higher under AHCI than IDE. Is this even possible?

 

[razel]

Yes. 4K Queue Depth of 32 is a specific benchmark to test if NCQ is working. NCQ is the main advantage with ACHI. With IDE your 4k and 4k @ 32 should be almost the same.

In real life home usage, your Queue Depth with an SSD is usually between 0 and 1. Their access times are so low that they finish disk requests so fast your queue is usually closer to 0.

 

[Arg Clin]

Thanks. I suppose that clear up some things, but raises the big question, if the performance gains are worth the decrease in write speed.

I just did the same thing on the E-350 / Crucial M4 - here the picture was much clearer. 50-100 % performance gains on read, and consistent 15-50 % gain on write.

 

[razel]

Don't worry about that dip in write speeds. The SSD needed to heal itself. The only real way to get a consistent benchmark test with SSDs to ensure that it will always work on an area that it doesn't have to do garbage collection on... which is an unrealistic result.

 

[Arg Clin]

Thanks