The Intel X25-M 80GB Solid State Drive Review

[Apoptosis]

Solid State Drives (SSD) have been around for over 25 years, but only recently have SSDs started to prove ready for the mainstream computing market now that performance is getting better and the price is coming down. It's been no secret that Intel has been working on a SSD for some time now and today they have finally released the new drives! Read on to see the X25-M in action!



Article Title: The Intel X25-M 80GB Solid State Drive Review
Article URL: http://legitreviews.com/article/788/1/
Pricing: $595 at time of print

[Gomeler]

Hot damn that's fast. I need to find when the next shipment is arriving near Atlanta so I can "borrow" half a dozen

[martini161]

dang, im going to be faced with a decision soon! i7 or this...

[Alathald]

I always thought SLC was the faster of the two and MLC was cheaper to produce ...don't really understand all this SSD stuff completely I'm afraid but I do understand that those are some very fast drives.

[Apoptosis]

You are correct! SLC is faster and is being used on the Intel Extreme SSD drives... MLC is being used on this one and is cheaper to produce because you get 2 bits for the price of one in a nut shell.

[Alathald]

Wow, in that case...just wow. Can't wait to see the speeds on the Intel SLC versions.

[Bwall]

More tests at 80% capacity please?! I'm really curious to see how this impacts multi-tasking performance.

[DMB2000uk]

I hate to think how expensive the SLC version is going to be.

I don't think I would want to pay quite as much for this drive knowing that the write speeds are less than what I currently have in my raid0 setup.

I'd love to have the silence and reduced heat though.

Dan

[dicecca112]

I haven't used the Intels, and probably won't because the price is far too high, but for MLC this is incredible. Kudos to Intel for creating an incredible controller. As for the prices people are complaining about, you have to remember that Intel is selling theses for more because they can, they beat anything on the market, including most SLC drives. I expected actually higher prices

[kenc51]

I haven't used the Intels, and probably won't because the price is far too high, but for MLC this is incredible. Kudos to Intel for creating an incredible controller. As for the prices people are complaining about, you have to remember that Intel is selling theses for more because they can, they beat anything on the market, including most SLC drives. I expected actually higher prices

Have you read the review over on AT?
Intel's MLC drives are faster as they have a better controller.
It seems current MLC drives (with a jmicron) controller have "issues" with write performance when stressed

[gwolfman]

Very nice review!

Btw, I noticed that you didn't post the "64MB File Test" for the VelociRaptors (VRs) in RAID 0. Think you can post that for comparison?

There are three things that get me:

• The write speeds on the Intel SSD are pretty bad in comparison to the VRs.
• Does the SLC-based enterprise version use the same or similar cache/buffer method? If so, why would any enterprise-level array enable the cache? They'd probably/should use the write-through method to avoid data corruption and I wonder how that would affect performance.
• I'm intrigued at the 80% full test. It's very interesting to see that. I never would have thought about that. So the "fuller" your SSD, the slower it goes. Does this happen with other SSDs like the MLC-based Samsung and the OCZ Core series?

Thanks!

[Apoptosis]

Post by gwolfman on 09 Sep 2008 09:32 am

Gwolfman,

I added the HD Tune Pro 64MB File test for the VelociRaptors in RAID 0 for you on page 4 - http://www.legitreviews.com/article/788/4/
I also added Sandra 2009 Read performance numbers that I did and forgot to post yesterday - http://www.legitreviews.com/article/788/7/
The last change I made to the article is that I added the ATTO Disk Benchmark I/O Comparison numbers, I did the test just didn't originally include them - http://www.legitreviews.com/article/788/8/

Hope those additions help add something to the article!

I'm working on the 80% testing as we speak, so expect an update on thursday/friday.

[gwolfman]

Post by gwolfman on 09 Sep 2008 09:32 am

Gwolfman,

I added the HD Tune Pro 64MB File test for the VelociRaptors in RAID 0 for you on page 4 - http://www.legitreviews.com/article/788/4/
I also added Sandra 2009 Read performance numbers that I did and forgot to post yesterday - http://www.legitreviews.com/article/788/7/
The last change I made to the article is that I added the ATTO Disk Benchmark I/O Comparison numbers, I did the test just didn't originally include them - http://www.legitreviews.com/article/788/8/

Hope those additions help add something to the article!

I'm working on the 80% testing as we speak, so expect an update on thursday/friday.

Thanks!!! I give you props for heavily interacting with your forum members and are willing to add requested info/tests. I'm starting to like this site because of that.

[gwolfman]

I added the HD Tune Pro 64MB File test for the VelociRaptors in RAID 0 for you on page 4
http://www.legitreviews.com/article/788/4/

For those following links, I think you meant page 5:
http://www.legitreviews.com/article/788/5/
Thanks again!

[Vanhelm]

Hello there!

Nice review of the impressive Intel X25-M Drive. I may have some insight on why the drive performs slower when at 80% capacity. I was reading the X25-M datasheet and found something interesting.

First of all, the datasheet can be found here.

For all that feel lazy to read everything there I'll just point out that in section 3.5.4 - Minimum Useful Life it states the following:

A typical client usage of 20 GB writes per day is assumed. Should the host system attempt to exceed 20 GB writes per day by a large margin for an extended period, the
drive will enable the endurance management feature to adjust write performance. By efficiently managing performance, this feature enables the device to have, at a
minimum, a five year useful life. Under normal operation conditions, the drive will not invoke this feature.

So, it seems to me that the decrease in performance is that this feature is being enabled on the drive, due to exceeding a 20GB a day writes. Remember, depending on each individual user, the drive usage would be different, but I doubt everyone will be exceeding the 20GB/day mark for an extended period of time. I just wanted to note what I found and maybe give some light as to why the drive is performing slower. I'll be on the lookout on what other findings you make about this drive.

[gwolfman]

^^^^ but it's not like DRAM. You don't have to rewrite every bit of data you read. I don't see how this "endurance" feature should affect read results. Possibly write results, but not reads. I didn't read the datasheet, but that's how it seems to me from the excerpt you posted.

[Vanhelm]

I see your point, and agree with you that read performance shouldn't be affected. Maybe Apoptosis can contact an Intel rep to look into that. Maybe it's a bug in the firmware...

[Apoptosis]

I see your point, and agree with you that read performance shouldn't be affected. Maybe Apoptosis can contact an Intel rep to look into that. Maybe it's a bug in the firmware...

The firmware has been updated since I was shipped the drive as I did the upgrade myself, so I know they have been making improvements over the past several months. From my understanding the firmware that I used for the benchmarking is the shipping version that will hit the streets.

[gwolfman]

I see your point, and agree with you that read performance shouldn't be affected. Maybe Apoptosis can contact an Intel rep to look into that. Maybe it's a bug in the firmware...

The firmware has been updated since I was shipped the drive as I did the upgrade myself, so I know they have been making improvements over the past several months. From my understanding the firmware that I used for the benchmarking is the shipping version that will hit the streets.

Interesting. Do you have to connect the SSD in any special way to flash it (like the mini-USB on OCZ Core v2) or does it just do it over the SATA data port it's already connected to?

[Apoptosis]

it was done over the sata connection... here is a before and after update, notice the version numbers.I did all sorts of testing on this drive, but only 30% of it made it into the article due to time constraints and the fact I had to do it twice with the firmware update.