6_6_6

Intel X25-M Impressions

Recommended Posts

CONCLUSION: No benefits to power user. Or not what you would expect.

I don't care about the IOPs or whatever tests people run at this, but do not fall under the false pretense that 11ms versus 0.1 ms make a worlds of difference. IT DOES NOT. As opposed to my Seagate 7200.11 750GB (more than 1 year old), i see 20-30% improvement which would probably be obtained by using a newer mechanical drive.

Intel X25-M 80GB (partitioned as 50GB for garbage collection crap)

Gigabyte X38-DQ6, Q6600

Intel AHCI Driver NCQ on

2008 Server x64

BOOT TIME

Seagate: 80 sec

Intel: 44 sec

Big deal. I boot my system once in 3 months.

PHOTOSHOP

Seagate: 16 sec

Intel: 11 sec

Zipping Windows Folder 5.41GB, 40K files

(on another partition at the same drive, all files read, appropriate permissions granted)

(also note that Seagate was heavily fragmented and the empty space was near the end of the drive)

Seagate: 22 min 30 sec

Intel: 20 min 10 sec

ANTIVIRUS

Seagate:

----------- SCAN SUMMARY -----------

Scanned directories: 946

Scanned files: 6211

Infected files: 0

Data scanned: 526.68 MB

Time: 106.377 sec (1 m 46 s)

--------------------------------------

INTEL

----------- SCAN SUMMARY -----------

Scanned directories: 946

Scanned files: 6211

Infected files: 0

Data scanned: 526.68 MB

Time: 104.552 sec (1 m 44 s)

--------------------------------------

Microsoft AHCI driver limits drive speed to 4 MB/sec while the drive reads/write 240/70 MB/sec as advertised with Intel AHCI driver.

Yes VMware opened up slightly faster and things snapped up faster at the beginning but this is not a big deal since it all gets pulled from memory later on.

NOISE: It is not silent. It is essentially the same as my Seagate because I have to run my Seagate along with this 80GB crap.

Overall, big freaking disappointment.

Edited by 6_6_6

Share this post


Link to post
Share on other sites

Do you have an insatiable urge to talk without exhibiting a modicum of logic or are you just getting paid by each post?

:) at you calling yourself a poweruser.

If all you do is some zipping and virus scans, maybe you should just order whats right now cheapest @ dell...

Share this post


Link to post
Share on other sites

There's something wrong with your driver / OS config. Also, your virus scan is probably CPU limited or something (not sure there, that doesn't jive with results I've seen myself).

The big test isn't how long it takes to virus scan, but how badly a virus scan or other background tasks (large directory copies) interfere with other work.

Another big test is loading large video games -- load times in levels go way down (and unlike boot times, this DOES matter), and any games with enough textures to texture-swap gain framerate too by stuttering less. I don't have one of these on my windows box but I can surely tell you that the hard drive gets in the way of my gaming experience and I'm eyeballing the price drop on these things.

The next big test is if you run enough apps to cause any sort of disk-swapping. Even with 4GB or 8GB a power user can cause paging and page-back on a traditional drive goes at random i/o rates near a handful of MB/sec while on this SSD go at 50x that rate.

As a developer pulling code branches (usually two full dev environments each on a different branch), compiling, running a db, running a VM, 400MB+ of firefox tabs, 10+ shells open with very long/large histories (1GB RAM at times), maybe a copy of GIMP, building and launching the apps under development over and over and reinitializing the db -- this thing makes a big difference.

Its comparable to having a lot more RAM -- it makes things a lot smoother, but like RAM if you don't need it it won't help much. But it goes hand in hand with Lots of RAM because it helps where the RAM can't -- when things have to go to disk and aren't just cached because you're doing complicated tasks.

Note, I have found that it is vital on Linux / MaxOS to do some file system tuning to make it perform best.

On Linux, use the noop scheduler and turn dirty_background_ratio down to the equivalent of 100MB or less (a setting of 1 for larger systems, google it if you don't know about it).

I have no experience tuning it for Windows, but if there are defaults in Linux that schedule I/O assuming its made from spinning platters, I'm sure Windows does similar.

There is no chance that the newer hard drives are 20% faster than your 7200.11. They are up to that much for sequential transfer only, and 0% faster for random access unless you get a 15k rpm SAS drive.

Your comment on it not being silent is silly. Might as well say your TV's mute button doesn't provide silence because you still have to breathe and pump blood during commercials.

Your copy test from another drive is clearly bottlenecked by the source drive, what did you expect? It to make the other drive faster too? What about other tasks concurrent with this? Did you open up PerfMon and look at the OS disk % utilization during the test? Try a directory copy within the same drive or something. Maybe a windows file search for something that can't be indexed forcing it to scan the drive? The biggest speed gains are concurrent activity and pure random access.

I agree that for many tasks it doesn't help much. Other drives and CPU bottleneck tons of stuff. Your use cases above just don't seem to be that demanding. Other SSD's would actually lead to slowdowns in a lot of things -- the Intel drive doesn't have that problem. Boot time tests are really silly for most people. They are less silly for a MacOSX user though since every freaking update from Mr. Jobs forces a reboot (even just a quicktime or itunes update -- its seriously like Windows NT 4 reboot-land).

I have also never bothered to partition my X25-M's under the full 80GB. I have seen the slowdowns on occasion (writes temporarily drop to 40-50MB/sec) but they have been temporary, and only affect heavy write scenarios. In fact I've only ever seen it during or just after benchmarking. It doesn't sound like you are a heavy-write user so there is no point in limiting it to 50GB.

The drive is already internally over-provisioned to some extent anyhow.

Share this post


Link to post
Share on other sites

There is nothing wrong with my config, please do not assume things out of the thin air. This is a clean system. Nothing runs at the background. Nothing is CPU limited on a Q6600 @3.2 GHz or its default speed.

This drive is fine. It matches review data. See graph below.

This drive just makes no difference in nothing to me. I have 8GB RAM, everything loads from RAM after initial startup. And when you run the system for 3 months without a reboot, there is nothing that is pulled out of the drive any more. And on basic operations like zipping, raring directories, this drive does worse than my Seagate.

There's something wrong with your driver / OS config. Also, your virus scan is probably CPU limited or something (not sure there, that doesn't jive with results I've seen myself).
The next big test is if you run enough apps to cause any sort of disk-swapping. Even with 4GB or 8GB a power user can cause paging and page-back on a traditional drive goes at random i/o rates near a handful of MB/sec while on this SSD go at 50x that rate.

I have no swap or pagefile. I have nothing swapping.

As a developer pulling code branches (usually two full dev environments each on a different branch), compiling, running a db, running a VM, 400MB+ of firefox tabs, 10+ shells open with very long/large histories (1GB RAM at times), maybe a copy of GIMP, building and launching the apps under development over and over and reinitializing the db -- this thing makes a big difference.

It makes absolutely no difference to me when i am writing mails in Eudora, processing network traffic real time with Ethereal (wireshark), running a virus scan, zipping a directory with PPMD, sawmilling 5GB bzip files for web reports, having 3 tabs open with IE7 and encoding with WMEncoder. The only difference is that I have no bloody space left for anything. Everything functions the same as it was with Seagate less space.

Your comment on it not being silent is silly. Might as well say your TV's mute button doesn't provide silence because you still have to breathe and pump blood during commercials.

Don't be ridiculous. Anyone using these drives are using them as OS drives and are bound to run additional drives concurrently. If you are getting everything done on a 70GB drive, you probably are dealing with 10K text files all your life and it wouldn't matter for you if you run an SSD or 10GB Quantum from 10 years ago for both would get your work done with no difference to you.

Your copy test from another drive is clearly bottlenecked by the source drive, what did you expect? It to make the other drive faster too?

Zip/virus tests done on these drives when they were not boot drives. Nothing has been copied/read/written to them from another drive and I dont understand how you came to this conclusion. I would like you to take a Vista windows folder and zip it (using legacy-default compression) and report here the time you will get. Or just take a 6GB folder with 40K files.

Yes, there is no stuttering or delays like others experienced with other drives (gosh, i dont even want to imagine that) but this drive is not any better for me than my Seagate with properly working NCQ for desktop use. It might be better for someone who runs his computer for an hour a day... or who is upgrading from PII 333MHz... Otherwise, waste of money.

x25-m.jpg

Edited by 6_6_6

Share this post


Link to post
Share on other sites

I also must add that the drive does not deliver constant throughput with random reads. If you are running it sequentially (say 3 copies of HDSpeed at different drive positions reading sequentially), it will deliver 240 MB/sec constantly. But as soon as you do drive operations that require accessing random data at random locations, it will not supply constant throughput. Throughput will fluctuate (measured by XPerf from Microsoft and also Task Manager > Resource Monitor).

Whereas in Seagate 7200.11, I get the same throughput whichever data i am reading at whatever location.

Share this post


Link to post
Share on other sites

Hi there,

There is currently only one situation where I would unconditionally recommend the X25m. If you have a laptop, and that laptop is your life (it's your workhorse), get an x25m. There are cheaper SSDs on the market. Ignore them for now (they may improve in the next 6 months). Get an x25m.

BBH

Share this post


Link to post
Share on other sites

BBH,

I would agree with that. It would be worthwhile even if it was for just to avoid the seek noise on laptops.

I also noticed that performance of the drive dropped since the 2 weeks I am using it even though it was secure erased and partitioned 70% capacity. In the next few days I will boot another drive and benchmark it again. This must be a joke.

I would say best upgrade for a power user is properly working NCQ implementation. This SSD hoppala is snake-oil for desktops. No wonder drive manufacturers did not see it as a threat.

Edited by 6_6_6

Share this post


Link to post
Share on other sites
I have no swap or pagefile. I have nothing swapping.

Ah, here is your problem, removing the pagefile will reduce performance.

Share this post


Link to post
Share on other sites
I have no swap or pagefile. I have nothing swapping.

Ah, here is your problem, removing the pagefile will reduce performance.

Damn. Did you have to say that? Now he's going to rant about pagefiles again.

Share this post


Link to post
Share on other sites

What the hell are you talking about? There is a seperate topic about pagefiles, go post there your toilet ramblings if you want and do not clutter my topic.

I have no swap or pagefile. I have nothing swapping.

Ah, here is your problem, removing the pagefile will reduce performance.

Share this post


Link to post
Share on other sites
Note, I have found that it is vital on Linux / MaxOS to do some file system tuning to make it perform best.

On Linux, use the noop scheduler and turn dirty_background_ratio down to the equivalent of 100MB or less (a setting of 1 for larger systems, google it if you don't know about it).

I have no experience tuning it for Windows, but if there are defaults in Linux that schedule I/O assuming its made from spinning platters, I'm sure Windows does similar.

On the contrary, linux does not align cylinder boundaries for SSD erase blocks. If you are using an SSD in linux, you are already having a poorly performing system.

Here is an exceprt from a recent article from Theodore Tso, maintainer of ext2-3 fses:

Full article here: http://ldn.linuxfoundation.org/blog-entry/...rase-block-size

I recently purchased a new toy, an Intel X25-M SSD, and when I was setting it up initially, I decided I wanted to make sure the file system was aligned on an erase block boundary. This is a generally considered to be a Very Good Thing to do for most SSD’s available today, although there’s some question about how important this really is for Intel SSD’s — more on that in a moment.

It turns out this is much more difficult than you might first think — most of Linux’s storage stack is not set up well to worry about alignment of partitions and logical volumes. This is surprising, because it’s useful for many things other than just SSD’s. This kind of alignment is important if you are using any kind of hardware or software RAID, for example, especially RAID 5, because if writes are done on stripe boundaries, it can avoid a read-modify-write overhead. In addition, the hard drive industry is planning on moving to 4096 byte sectors instead of the way-too-small 512 byte sectors at some point in the future. Linux’s default partition geometry of 255 heads and 63 sectors/track means that there are 16065 (512 byte) sectors per cylinder. The initial round of 4k sector disks will emulate 512 byte disks, but if the partitions are not 4k aligned, then the disk will end up doing a read/modify/write on two internal 4k sectors for each singleton 4k file system write, and that would be unfortunate.

Vista has already started working around this problem, since it uses a default partitioning geometry of 240 heads and 63 sectors/track. This results in a cylinder boundary which is divisible by 8, and so the partitions (with the exception of the first, which is still misaligned unless you play some additional tricks) are 4k aligned. So this is one place where Vista is ahead of Linux…. unfortunately the default 255 heads and 63 sectors is hard coded in many places in the kernel, in the SCSI stack, and in various partitioning programs; so fixing this will require changes in many places.

Share this post


Link to post
Share on other sites
Intel X25-M 80GB (partitioned as 50GB for garbage collection crap)

Gigabyte X38-DQ6, Q6600

Intel AHCI Driver NCQ on

2008 Server x64

Did you do anything to allow for NTFS cluster size of 32 or 64KB or did you keep the 4KB cluster size default ?

Although, Win2008 Performance tuning guidelines can be usefull (cf NumberOfRequests io reg param)

Share this post


Link to post
Share on other sites

I kept 4KB default. However, I tested with aligned partitions and saw no tangible benefit. Changing default cluster size renders most drive utilities useless, so it is running with a default 2008 install with default cluster size.

If anyone requires any tests, i would be happy to run it for them.

Did you do anything to allow for NTFS cluster size of 32 or 64KB or did you keep the 4KB cluster size default ?

Although, Win2008 Performance tuning guidelines can be usefull (cf NumberOfRequests io reg param)

Share this post


Link to post
Share on other sites

sasfan, I am interested in a desktop replacement, not a DB server but I will include the results for your reference.

100% Random Write (Intel boot drive, Seagate secondary. No disk activity except minor OS read-writes, Superfetch, Prefetch, Indexing, etc disabled)

8K

Intel: 3600 IOPs 28 MB/s

Seagate: 242 IOPs 2 MB/s

Intel: 5000 IOPs 40 MB/s (Queue Depth 3)

Seagate: 710 IOPs 6 MB/s (Queue Depth 3)

256K

Intel: 150 IOPs 39 MB/s

Seagate: 432 IOPs 108 MB/s

Intel: 135 IOPs 33 MB/s (Queue Depth 3)

Seagate: 260 IOPs 65 MB/s (Queue Depth 3)

There is no steady throughput on Intel drive (be it OS or secondary, numbers always fluctuate) whereas Seagate delivers a constant stream.

IOPS of 100% 8k Blocks random writes @ NTFS with default cluster size would be interesting (e.g. used by Oracle databases @ default).

Share this post


Link to post
Share on other sites

Yeah, you can see that the controller is happy with small random writes, or sequential writes. But when you do large random writes, it's all over. I wanted to use an X25 (either M or E) for 1MB block reads/writes, but the writes ended up dragging perf way way down.

Well, hopefully someday soon they'll have more consistent (and good and predictable) behavior.

Share this post


Link to post
Share on other sites

Oh my goodness! I havent seen something like this.

My friend brought over his OCZ Core and you can't do two tasks at a time without having long long pauses in the system! It is like running a computer circa 1990s -- on floppies!

I can't believe there are actually people who pay money for this OCZ crap. If OCZ paid me money, I wouldn't use one. Yet he is happy. He says it is better than his 60GB Samsung. I would not wish this torment for my worst enemy.

My Intel apparently is godsent! I will stop complaining and sell this lesser crap to someone else.

Share this post


Link to post
Share on other sites

On the OCZ forum, there is a lot of discussion of the benefits of free space consolidation for improving write speeds for their SSDs.

http://www.ocztechnologyforum.com/forum/sh...ead.php?t=52329

Diskeeper shows this comparison for their Hyperfast software:

http://theovalich.wordpress.com/2009/01/28...zation-utility/

But this indicates that the X25-M is much more resistant to that type of slowdown, though it seems to be dependent on which firmware you have:

http://www.diskeeperblog.com/archives/2008...fast_is_al.html

Have you tried free space consolidation? It won't do anything for reads, but could help the writes.

Share this post


Link to post
Share on other sites

Don't even compare the OCZ Core... if you're looking at any OCZ drives, it needs to be the Vertex or the Samsung SLC based one...

NOISE: It is not silent. It is essentially the same as my Seagate because I have to run my Seagate along with this 80GB crap.
Don't blame the X25-M for that one then... :P

Share this post


Link to post
Share on other sites
On the OCZ forum, there is a lot of discussion of the benefits of free space consolidation for improving write speeds for their SSDs.

jf2000 thanks. Yeah, there is a lot of discussion about improving write speeds... There is also a lot of discussions about spending your days and nights stuck at boot screens, messed up boot-sectors, unfunctional drives and all.

No thanks, I think i will pass. OCZ is a piece of UTTER GARBAGE. I never experienced something so disasterous in my life. This is CRIMINAL. I feel sorry for anyone who has one. My friend will bring his system later on to do some tests... But honestly, I do not see no need for testing anything on this garbage.

Diskeeper shows this comparison for their Hyperfast software:

http://theovalich.wordpress.com/2009/01/28...zation-utility/

That pretty much is diskeeper's own advertising. It hasn't been verified by anyone outside. To the contrary, no one saw any benefits for Intel x25-M.

Please let's keep in mind that I am trying to get a replacement drive for my system. My Intel X25-M performs as advertised.

HOWEVER, I SEE NO BENEFITS FROM A USER'S DESKTOP EXPERIENCE PERSPECTIVE. IT IS NOT WHAT I THOUGHT IT WOULD BE. WHEN YOU ARE LOOKING AT 100X IMPROVEMENTS IN ACCESS TIMES, YOU ARE AT LEAST EXPECTING SOME PERFORMANCE GAINS. NOT HALF-ASS-BAKED 10% HERE, 30% THERE.

I DID NOT SEE HARD DRIVES COMING OUT AND SAYING, "Oh, we made floppy drives 30% faster than existing floppies. But instead of having 1.44MB, you will be having 0.1MB and you will have to stack 15 of us to be able to compensate for your existing setup".

I ALSO DO NOT HAVE TO SPEND MY DAYS AND NIGHTS TWEAKING THEIR CRAPPY DRIVES TO MAKE THEM WORK.

PS: I don't play games, haven't tested one.

Share this post


Link to post
Share on other sites
Don't even compare the OCZ Core... if you're looking at any OCZ drives, it needs to be the Vertex or the Samsung SLC based one...

In Anand's tests, I don't see them coming halfway close to Intel. They might have fixed stuttering issue, but they are still bigger piece of crap nonetheless. OCZ has been selling these POSs since 2 years and now they managed to fix the most important issue?

Don't blame the X25-M for that one then... :P

Why not? I never ran 2 drives at the same time in my life due to noise concerns. Now I have to. I don't have to run an additional drive with any of my other drives. Only with Intel. And I also must mention that I have been waiting for these SSDs close to 2 years. I was even about to buy an OCZ one when it first came out.

Share this post


Link to post
Share on other sites
I kept 4KB default. However, I tested with aligned partitions and saw no tangible benefit. Changing default cluster size renders most drive utilities useless, so it is running with a default 2008 install with default cluster size.

If anyone requires any tests, i would be happy to run it for them.

I would like you to try running a test with a 64KB NTFS cluster size (It's aligned at 1MB boundary in Win2008 which is ok for me).

I think there is a real performance point out of this "page size".

Of course, it may be disappointing to have to modify some Windows defaults...and it is obviously a very very very bad idea regarding the SSD desktop market feedback.

NB: I hope you did not disabled the "write cache" !

Share this post


Link to post
Share on other sites
I DID NOT SEE HARD DRIVES COMING OUT AND SAYING, "Oh, we made floppy drives 30% faster than existing floppies. But instead of having 1.44MB, you will be having 0.1MB and you will have to stack 15 of us to be able to compensate for your existing setup".

I ALSO DO NOT HAVE TO SPEND MY DAYS AND NIGHTS TWEAKING THEIR CRAPPY DRIVES TO MAKE THEM WORK.

Every hard drive in existence in the history of the world needs to be defragmented for optimum performance. With SSDs, it is even simpler, since you only need to do free-space consolidation. But at least you have clarified the limits of your capabilities.

Share this post


Link to post
Share on other sites

Please do not spread bogus information.

Defragging an SSD is not only useless but it is harmful to the life of the drive and to its performance. OS or defrag utilities have no way of knowing what space is free and what space is not since sector allocation is being done by the drive's firmware.

Every hard drive in existence in the history of the world needs to be defragmented for optimum performance. With SSDs, it is even simpler, since you only need to do free-space consolidation. But at least you have clarified the limits of your capabilities.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now