PcPer Analyzes Indilinx Fragmenting

alcoholbob

Diamond Member
May 24, 2005
6,364
425
126
PCPer has an obsession with file creation speed. I don't know about you but creating large files seems rather uncharacteristic of normal use, especially on small hard drives. Most wear leveling will cause a drop in sequential speeds due to increasing fragmentation but random read/writes don't get affected past steady-state.

And to me, the Intel firmware was simply offered for PR purposes. The X25-E degrades similarly and the wear leveling makes zero difference from normal desktop use to professional enterprise/server use. In fact, the X25-E went has gone through production firmware revisions (8621 to 8790, and an 85xx series before that) and none of them "fixed" the performance degradation that the new X25-M firmware does. In other words, it's just to to plug in some PR hole (the reviewers and their obsession with sequential benchmarking).

I think TRIM is for people who are obsessed with benchmarking performance.
 

n7

Elite Member
Jan 4, 2004
21,281
4
81
I haven't read the article yet (will when i get home), but my X25-M was stuck @ 30 MB/s write for ANYTHING.
Big or small writes, it couldn't do over 30 MB/s.

I installed the firmware update & it was back to normal write speeds...not sure how that's PR only :confused:

If i misunderstood what you were saying, nevermind.
 

shabby

Diamond Member
Oct 9, 1999
5,782
45
91
Even the $1300 vertex EX has the same problem, throwing expensive memory on a so-so controller simply wont fix it, no idea why ocz even released that drive.
Glad i spent the cash on the x25-m.
 

alcoholbob

Diamond Member
May 24, 2005
6,364
425
126
Probably not relevant for the sequential benchmarks but the 64-thread 4K numbers indicate the reviewer was running in IDE mode. In AHCI it will go as high as 80MB/s. The same is true of the vanilla Vertex--about 25MB/s in 64-thread reads, 60MB/s in AHCI.

Of course nothing compares to Intel's 150MB/s.