Crystal Disk Mark?

[Santilli]

http://crystalmark.info/software/CrystalDiskMark/index-e.html

Anyone used this? Opinions?
Observations?

Seems a new benchmark is needed for SSD's.

 

[ddrueding]

Seems interesting. Looks to measure the stuff that has been determined to matter in SSDs. Whether it measures it well is another matter. Here is a run from my laptop while it was in the middle of a bunch of other stuff.

 

[ddrueding]

Just ran it a number of times after a fresh reboot. Scores were +/-10% with no trends...not a good sign.

 

[Santilli]

HMMMM.

I thought nearly all laptops shipped with SATA 1?
Why is it over 150 MB/sec?

 

[ddrueding]

Because it is a really nice laptop

 

[Mercutio]

HMMMM.

I thought nearly all laptops shipped with SATA 1?
Why is it over 150 MB/sec?

I suspect that got better with the advent of modern notebook chipsets. He has a Thinkpad T510, after all.

 

[ddrueding]

410, actually, on your recommendation. Couldn't be happier.

 

[Howell]

The T60 supports sata II as well.

 

[Santilli]

Not a bad setup. Thanks David;-)

 

[LiamC]

Just ran it a number of times after a fresh reboot. Scores were +/-10% with no trends...not a good sign.

Why isn't it a good sign? Those numbers look very good to me for an SSD.

 

[ddrueding]

The numbers are OK, but if it can't manage better than a 20% swing on consecutive runs, I question their methodology.

 

[LunarMist]

Results are rather pointless with SSD. I have no idea what they are getting at and don't care.

 

[Santilli]

Results are rather pointless with SSD. I have no idea what they are getting at and don't care.

HMMM. There are some REALLY bad SSD's out there. Nice to have a bench mark to find out why they are. I have one that won't clear 30 MB/sec. Makes a passable USB 2 external drive, but that's about it. Waste of 140 dollars...

 

[LunarMist]

HMMM. There are some REALLY bad SSD's out there. Nice to have a bench mark to find out why they are. I have one that won't clear 30 MB/sec. Makes a passable USB 2 external drive, but that's about it. Waste of 140 dollars...

According to the crystal dickmark the write speeds are all over the place, sometimes half of what they are otherwise. :viking: Maybe I should throw out the X25-E.

 

[Santilli]

According to the crystal dickmark the write speeds are all over the place, sometimes half of what they are otherwise. :viking: Maybe I should throw out the X25-E.

OK:

So, we now know CDM doesn't work for your drive, in your system. Does make me wonder if the limitation is somewhere in your system.

I know the SATA ports on this Gigabyte board seem to be 4 VERY fast, 6 seem to be SATA 1, with actually only 8 of the 10 really working.

 

[Mercutio]

ddrueding doesn't seem to think it's working either. Maybe it's just not a good tool.

 

[LunarMist]

The crystal dickmark results are inconsistent with three other apps, the manufacturer's specs, and widely reported results.

 

[Mercutio]

I feel like I should say something crass about dickmarks, but I can't think of anything that is not also completely gross.

 

[sechs]

Is this good?

 

[LunarMist]

It is pink and purple, so it must be good.

 

[Santilli]

Small Ram disk from the beast:

 

[LiamC]

The numbers are OK, but if it can't manage better than a 20% swing on consecutive runs, I question their methodology.

If any benchmark swings 20% or more between runs, I questions that benchmarks methodology, because it's telling you precisely jack... That isn't a benchmark, that's a lottery.

 

[ddrueding]

RevoDrive 180GB

 

[ddrueding]

Some more results from my home machine.