Intel X-25 V vs Intel 320 120GB

[mlaidman]

Hi folks.
I just installed an Intel 320 120GB drive as the X-25 V 40GB budget boot drive from a few years ago was starting to push it for space...
I've run HD Tune Pro (the trial version) a few times, although I could only get the Read test to run...The Write test said I must remove all other partitions first. I've only got one Windows partition, but all my other SATA ports are filled with other drives, so I'm not sure why it was saying that.
Anyways, I'm curious about the results:

X-25 V: (rough averages)
Min: 162
Max: 220
Avg: 202
Access Time: 0.068
Burst: 103

320 120Gb: (rough averages)
Min: 89
Max: 208
Avg: 192
Access Time: 0.080
Burst: 167

This was on a fresh install of Windows 7 Professional, Gigabyte MA770t-UD3P board with the latest firmware, 8GB RAM, X3 720 CPU, Radeon 6850...
Both drives had the latest firmware, and I'm running AHCI...

Any thoughts on why I'm getting lower speeds (especially the min and averages!) with the newer drive? Especially as it's a higher capacity unit? The slowest mid-80s bit is reached in the very first part of the drive, before the "12Gb" marker at the bottom of the HD Tune window.

Then again, this is ONLY the read test of ONLY ONE benchmark program, but it still seems like there should be an improvement.

Are bum SSD drives all that common?

Thanks

 

[BFG10K]

I have the same drive and I get similar HD Tune read speeds. I don’t think HD Tune accurately measures sequential read speeds on SSDs, though write speeds seem to be accurate. Try AIDA64 or CrystalDiskMark instead (I get over 270 MB/sec with Crystal).

Also HD Tune’s write tests destroy data so you can’t have any partitions. It’s actually best to run all HD Tune benchmarks with the drive in an un-partitioned state. That stops the OS and/or file system from accessing the drive while the tests run.

As for the 12 GB marker, I’m guessing that’s how much data you have on your partition. I only found this out recently, but SSD (and flash drive) read speeds are affected by whether there’s data on the drive.

If you run a sequential test across an entire SSD, the parts that have data will have slower read speeds than those that are empty. I get the same on my SSD across its two partitions. Where there's data the read speed drops, then it suddenly jumps back up exactly where the data ends.

And again, if you have the OS on the drive, it’s probably accessing the drive while the benchmark is being run so it causes those momentary dips.

 

[Coup27]

Post a crystaldiskmark score and it'll be a lot easier to do and easier for us to say if the figures are on spec or not. Just do one pass @ 1000MB. No need to thrash your SSD 5 x 5000MB.

 

[mlaidman]

Thanks for the replies.
This is what I get with CrystalDisk:

258.4 / 137.1
172.1 /139.2
18.04 / 28.73
140.3 / 61.02

My 4K writes still seem a little slower than the few other benchmarks I could find, but I did only run it once and everything else seems to be in the ballpark.

It's interesting what you said about reads being slower over the portion of the drive that's already full. I suppose it makes sense, though.

Thanks again folks. I guess everything looks OK, although I no longer have the 40GB drive connected to check what it does with CrystalDisk......need some more SATA ports......

 

[Coup27]

All of those figures are about right apart from the 4k reads. My X25-M G2 used to give me 25mb/sec but that was on an Intel system. Maybe the slightly slower numbers are due to it being an older AMD system. A lot of peoples benchmarks these days are on Intels Sandy Bridge.

 

[mlaidman]

Hmmm...that makes sense about the platform differences. So it looks like everything is within spec. That's good!
Thanks for your help!