something SSD you didn't know about from OCZ

[p2501]

Tony, did you also do this combination of a little benching and experience testing with an Vertex type drive? I'm contemplating to buy that 30 or 60GB Vertex just for my V64 install. What is it exactly that makes the Vertex go faster? Is it just the cache or maybe a better controller?!

[HDCHOPPER]

o ya ssd's are happening ! Apex is next for me 2 in raid

[m^2]

Tony,

thanks for the great info, my next build will definatly be a few SSD in raid 0

will you guys be making any smaller SSD maybe 15gb or so?
after 7 yrs or so of running 24/7, some of my 6gb HD are starting to show thier age in my crunchers in the basement. so ill need to start replacing them soon

Super Talent has them.

[LowRun]

PC perspective has a review of the Apex up: http://www.pcper.com/article.php?aid=661

I need to first explain some problems we had with our IOMeter testing and the new OCZ Apex SSD. The first returned results from this benchmark were absolutely abysmal for three of the four tests (the web server scenario for some reason was immune) and included IOs per second in the teens. Consider that the Intel drive was going over 14k IOs per second in the web server tests this was obviously an eyebrow raiser. We initially thought we were seeing an incompatibility with the Apex internal RAID-0 configuration and the way that IOMeter does its hardware level testing and that may still be the case. OCZ eventually found us a work-around by telling us to set the "starting disk sector" to something like 512 or 1024 and then set the "maximum disk size" to anything reasonable like 2,000,000.

That did improve our results and allow the Apex 250GB SSD to give us "normal" testing scores. That being said, there is something amiss with this application and this particular drive as no other HDD or SSD has exhibited any similar results. I am going to continue looking into the incompatibility with IOMeter in the coming weeks hoping to find the culprit and thus be sure there is nothing wrong with the JMicron-based internal RAID architecture design.

I'd say it's urgent to wait until the dust settles.

[One_Hertz]

2,000,000 is nowhere near reasonable for a maximum disk size setting. The IOPS in the teens is not an incorrect result, that is the performance of the drive under certain work loads and the cause of the stutters.

[RyderOCZ]

Low IOPS has nothing to do with stuttering.

[One_Hertz]

Low IOPS has nothing to do with stuttering.

I believe Tony said that these are on the same controller as the older V2 drives, so that gives it about seven (7) random write IOPS (I have V2 drives myself).

As far as I understand the drive ques up requests so if it is told to write to say 14 random sectors and then to read something, the read request has to wait for the writes to finish, which would cause a 2 second long "stutter" during which the drive finishes its writing procedures. So I guess IOPS are not the direct cause of the stutters, but if they were 100 instead of 7, the length of the stutter would be very tiny and would make it a lot more acceptable. I may not have things perfectly right though.

[chispy]

Great info here Tony , thanks for sharing. im looking forward to buy my next SSds and definitely will be OCZ SSds . I have my 4 x 30GB OCZ Core Series V2 Since their release and I have not seen a stutter on my raid0 Set up. Looking forward to more post like this Keep up the good work Tony.


chispy.

[NickK]

http://benchmarkreviews.com/index.ph...&limitstart=12

[orangekiwii]

BAH!!!! don't get into SSDs yet wait a couple more months all of the cheap ssd's mlc drives that is are terrible and conventional hard drives are better

wait for them to surpass conventional hdds and then get into the market but as of now its not so good O_O