Community
cancel
Showing results for 
Search instead for 
Did you mean: 
idata
Community Manager
4,093 Views

Intel Matrix Storage + SSD TRIM Pass-through?

Hello!

I would like to know if Intel is aware of the issues with SSD's and RAID arrays regarding the lack of TRIM capability. More and more SSD drives support the block cleaning/optimising method aka. TRIM. Windows 7 has the operating system level support for it. As far as the user useres his SSD's in normal SATA/Legacy IDE mode this TRIM methid works because the OS can send the TRIM method calls directly to the drive itself. But if the user's SSD drives are in a RAID0 array TRIM capabilities are gone, because the Intel RAID controller does not pass through the TRIM method calls.

My question is that whether Intel is aware of this problem and if is there any effort to solve this issue? Are we seeing any driver updates soon?

Thank you indavance!

Regards

0 Kudos
22 Replies
idata
Community Manager
137 Views

Jozsef,

I personally wouldn't be too concerned with the TRIM capability to begin with, especially in a RAID configuration. There won't be a whole lot of performance gain since these SSDs are already very fast. Aside from that, your RAID controller is where the performance enhancements should be looked at. Are you using a RAID controller with a built-in cache? If it's an on-board RAID controller, probably not, and you won't see much performance gain by using RAID on these controllers -- just redundancy gains. I currently have an Intel-based RAID ICH10R and under all configurations of RAID 0 or 1 with two X25-E drives, the performance didn't change much from using just a single drive. I ended up buying an Adaptec card (the 5045) to see true gains. Perhaps it was a combination of the chipset and motherboard, but unless you are writing hundreds of thousands of tiny files, I don't see where your benefit for this feature would amount to much.

Personally, I feel the TRIM command defeats the purpose of wear-leveling, sacrificing lifespan for very little performance gains (there have been exceptions). Maybe there's still a significant difference for the X25-M series, but I've never used those. Anybody have links to some more information regarding TRIM benefits on the M series with the updated firmware? Other brands of SSDs wouldn't fit this topic.

Hope this helps.

idata
Community Manager
137 Views

Thank you for your response but I think you might not get my point. I am not too muck into performance gains. The problem is that SSD write speeds degrade as you write more data on the drive. It is no concern with NAND chips wearout it is a matter of the method that SSD's use for writing and overwriting data. You can describe it as a kind of defragmentation but it is not that simple. You should check on this article if you are not sure what is TRIM: http://anandtech.com/storage/showdoc.aspx?i=3531&p=1.

If you do not get any performance increase with RAID0 you have obviously going wrong.

TRIM support is more likely to be solved on the RAID controller's side.

idata
Community Manager
137 Views

Hi Jozsef,

I actually do understand the issues of TRIM and SSDs fully. I've posted several times about this topic on past threads. However, I'll emphasize my point, too: The gains you get from using TRIM are minimal in real-world use. Sure, you can see slight differences by running benchmarks, but odds are the system will not operate this way in real-world conditions.

RAID benefits significantly by having a built-in cache, thus the difference between enterprise-level controllers and 'low-cost' home consumer controllers. Gaining RAID performance isn't just a matter of 'going wrong' but a matter of using the right combination of hardware. It's a proven science, not open to much theory. A RAID cache will benefit your write performance more than the TRIM will, and that's a fact of I/O, period. If you wish to wait for TRIM to be fully implemented, that's your choice, of course, but I will guarantee your wait will be rewarded with very little satisfaction, except perhaps on the cheapest of SSD components.

However, I only speak for the majority of systems operating out there. Perhaps you have a very unique situation where you're constantly writing gigabytes of data on an hourly basis to your SSDs. But, if it happens to be for your system pagefile or temporary files, perhaps an additional RAMDISK would be the next best upgrade to your system. See other posts on this page for those ideas.

Respectfully,

William

idata
Community Manager
137 Views

zulishk wrote:

Personally, I feel the TRIM command defeats the purpose of wear-leveling, sacrificing lifespan for very little performance gains (there have been exceptions).

 

I don't think you understand how TRIM works. It isn't sacrificing lifespan, it is substantially increasing it, by reducing the frequency of read/erase/write cycles being needed. This increased lifespan IMO is a bigger reason to use it than the performance gains. (As you said the SSD is already fast enough without it).

idata
Community Manager
137 Views

william, it seems you are unawares of the multitude of mlc drives that are available do experience performance degradation w/o the sort of "cleanup" trim offers. regarding raid, i certianly notice the speed difference between using a single drive and 4 in a raid0 array, and would really like to have discard/trim support on my intel based workstation ich (which uses system ram for cache, if i'm not mistaken), which by the way offers very good performance (up to and including x4 ssd) compared to the adaptec you stated you have (i also run one, and the ich option performance is quite close). mlc drives are a bit different than the slc, and can really benefit from trim/discard. even slc can experience similar performance degradation effects over time w/o it's implementation.

idata
Community Manager
137 Views

@m.oreilly

I am aware of what TRIM is, and why it exists, and where it's utilized, and the benefits it could potentially give. However, most people do not see the larger picture. They focus on this one aspect without regard to the whole picture, something which I find usually to be an excuse to just wait.

Let's say TRIM is fully supported. Let's assume wear-leveling is still being used. Let's also assume that, for whatever reason, the system is writing way more data than it reads. Let's also assume SSD is being used for speed and not mass storage. This should cover about 90% of the SSD users out there.

TRIM will only clear pages which have NO data in them. In a typical operating environment, after a period of weeks or months, how many pages do we think would have NO data written on them? Even if it's 1 byte of data, the page has to be re-written like all the others. Without any standard supported defragmentation API for the SSD to create contiguous blocks of data at the page level, how can we be certain there will ever be any empty pages of data again? What good is TRIM if there are no or few pages to actually clear? TRIM cannot do the job by itself.

So, a lot more has to happen before TRIM is remotely useful, but even still -- if TRIM is the most important aspect of the purchase, perhaps we're looking at the wrong solution. With 4 SSD drives in a proper RAID 0 configuration, the amount of performance degradation will be humanly unnoticable -- that is, unless you happen to be writing gigabytes of data every hour.

Most systems read 75-90% of the time, and write 25% of the time or less. Writing is the only times when TRIM would have any effect.

I to not claim to know everything about TRIM or SSDs, but at least I have an operational system and it's much faster than before. Enough that I'd never go back to HDD without a fight.

TRIM is simply not an excuse to wait.

Good luck!

William

idata
Community Manager
137 Views

william, i am 'waiting" for trim and all the goodies that are involved in the standard, but have been enjoying several raided ssd arrays for quite some time, so i don't get this "waiting" you are referring to and yes, i notice when my performance goes down. anything that helps with this is a good thing. at the moment, i have to either do an hdderase or destructively re flash my drives to "restore" them, as being in an array, i am unable to take advantage of the brute force methods of trim that a certain manufacturer has come up with (single drives only). i work and play w/my system, and the drives get used. it almost seems as if you are declaring trim/discard to be some sort of boondoggle.

idata
Community Manager
137 Views

@m.oreilly:

IMHO, TRIM itself is pretty much a boondoggle, or at the very least, not even close to the Holy Grail. Post some links to testing results which show improvement from TRIM in real-world situations -- and NOT from completely erasing a drive (as this is not what TRIM does). In real-world situations, the TRIM command would have very little to clean up without a proper internal defragmentation of the flash pages, and that would have to also need to include swapping data from one page to another to ensure proper wear-leveling.

As I mentioned, if you are writing gigabytes of data, you just might see stacked up speed improvements from a new or majorly "trimmed" drive. Don't forget, it's also possible that if you are seeing slow downs from everyday use, perhaps the SSD isn't the primary issue. Maybe it's RAM and the paging file, or the CPU, or some software? Maybe it's just flat out cheap SSD hardware? I have a strong feeling SSD speeds will quickly outpace any benefit TRIM could possibly hope to achieve, and prices are constantly diving.

Is your application a unique one? Please let me know about it so we can share and discuss the results. You said you gain improvement by wiping your drive completely, but it's not what TRIM will do. Therefore, how can you come to a final conclusion that TRIM will address your concern? You might want to take a closer look at what is doing so much writing to your SSDs -- and add on a hardware RAM drive to supplement it.

Good luck!

William

P.S. Do you have documentation to support the fact HDDerase or the other method you use is actually performing the erase function you're looking for? "SECURE ERASE" is not necessarily the same operation as TRIM. If secure erase is working, then perhaps just clearing slack space will also help in your configuration...?

idata
Community Manager
137 Views

this is an intel forum, so i'm not going to espouse another manufacturer's product/FW. suffice to say if you google a certain popular company's sata/flash support forum,you will notice that many patrons of that site are indeed recouping drive performance using a proprietary app and drive FW capable of carrying out trim commands. i know what trim doesn't do, and i don't know why you are not being more supportive of the standard which will allow improved performance and broader acceptance of ssds. william, it's time to move forward, there is nothing to be gained by the dragging of feet. enjoy your drives

idata
Community Manager
137 Views

I don't feel there's any reason to avoid posting links or referencing information which is applicable to this topic, even if it's from another company or product. Perhaps it needs to be pointed out to those who need to know.

We never know - Intel might follow up with a response which reads something like, "Our drives won't require 'trim' support because they already perform internal optimizations, or they will in the next firmware release."

idata
Community Manager
137 Views

Well gentlemen, 5 months have passed since you began your interesting discussion/debate regarding TRIM and RAID. Intel firmware updates and Windows 7 have since been released. Any new performance data or changed opinions? I'm planning on implementing RAID-1 (for redundancy not performance) probably using two Intel X25-M SSDSA2MH080G2 SSD's for a new Win 7 Pro 64bit system I'm building. Motherboard will be Intel DP55KG.

 

I was definitely planning to implement TRIM if the onboard RAID controller will pass it through, but now I'm not sure. Most daily writes (temp, paging, backups, etc, will be going to RAM drive and/or a 1TB WD Caviar Green or Black RAID-1 disk array. Low noise, reliability and energy efficiency are more important than performance for me although read performance from the SSD RAID-1 array should be phenomenal--especially for home use.
DZand
Valued Contributor I
137 Views

1. None of the "Intel(R) Matrix Storage Manager" (MSM) driver versions do support the Trim command.

2. I don't think, that Intel will ever release an MSM driver, which passes the Trim command, because the MSM drivers will be outdated very soon.

3. Obviously Intel is planning to release in the near future a completely new AHCI and RAID driver generation named "Intel(R) Rapid Storage Technology" (RST). The first drivers v9.5.0.1037 are already available at Station-Drivers. They are optimized for being used with SSD's and Win7.

4. What I don't know is, if the new RST drivers do already or will support Trim. That is why I started this thread:/message/74253# 74253 http://communities.intel.com/message/74253# 74253

5. The biggest problem for RAID users is the fact, that they are not able to run the Optimizer of Intel's SSD Toolbox.

idata
Community Manager
137 Views

Fernando, thanks for the info on MSM and RST driver support (or lack thereof) for TRIM and the SSD optimizer toolbox. Also read your thread (74253). Hope Intel responds.

Obviously I need to do some more research. Will the new RST drivers work on Intel's DP55KG motherboard? I'd never heard of Station-Drivers before. Looks like a French website. Do they have a reputation for getting drivers before everyone else?

DZand
Valued Contributor I
137 Views

steppinwolf schrieb:

Fernando, thanks for the info on MSM and RST driver support (or lack thereof) for TRIM and the SSD optimizer toolbox. Also read your thread (74253). Hope Intel responds.

Obviously I need to do some more research. Will the new RST drivers work on Intel's DP55KG motherboard? I'd never heard of Station-Drivers before. Looks like a French website. Do they have a reputation for getting drivers before everyone else?

The v9.5.0.1037 RST drivers are WHQL and are running perfect with my 2x160 GB Intel X25-M SSD's in RAID0. I got a much better performance than with the latest MSM drivers. For further details you may look into this thread, where I have shown my results with these drivers (I'm sorry, it's in German language): http://www.german-winlite.de/wbb/index.php?page=Thread&threadID=3141 http://www.german-winlite.de/wbb/index.php?page=Thread&threadID=3141

Since your mainboard has a P55 chipset and the RST drivers do support all Intel Southbridges from ICH7 up, you will have no problem to get them installed.

And yes, Station-Drivers very often publish drivers, which are shown by other companies several weeks later.

Regards

Fernando

PPapp
Novice
23 Views

Fernando wrote:

steppinwolf schrieb:

Fernando, thanks for the info on MSM and RST driver support (or lack thereof) for TRIM and the SSD optimizer toolbox. Also read your thread (74253). Hope Intel responds.

Obviously I need to do some more research. Will the new RST drivers work on Intel's DP55KG motherboard? I'd never heard of Station-Drivers before. Looks like a French website. Do they have a reputation for getting drivers before everyone else?

The v9.5.0.1037 RST drivers are WHQL and are running perfect with my 2x160 GB Intel X25-M SSD's in RAID0. I got a much better performance than with the latest MSM drivers. For further details you may look into this thread, where I have shown my results with these drivers (I'm sorry, it's in German language): http://www.german-winlite.de/wbb/index.php?page=Thread&threadID=3141 http://www.german-winlite.de/wbb/index.php?page=Thread&threadID=3141

Since your mainboard has a P55 chipset and the RST drivers do support all Intel Southbridges from ICH7 up, you will have no problem to get them installed.

And yes, Station-Drivers very often publish drivers, which are shown by other companies several weeks later.

Regards

Fernando

So are you saying that these RST drivers will pass TRIM through to a RAID array?

DZand
Valued Contributor I
23 Views

D749 schrieb:

So are you saying that these RST drivers will pass TRIM through to a RAID array?

I don't know, if the already available RST drivers v9.5.0.1037 do pass the TRIM command, but I expect, that we will see new RST drivers in the near future, which will support TRIM.

Since a lot of users with Intel Postville SSD's are interested in this question, I have startet this thread: /message/75320# 75320 http://communities.intel.com/message/75320# 75320, but I didn't yet get any answer from Intel.

Regards

Fernando

idata
Community Manager
137 Views

This link may not be too relevant to our discussion, but it's certainly good to know that there are also other issues out there worth looking into.

http://techreport.com/articles.x/16979/2 http://techreport.com/articles.x/16979/2

idata
Community Manager
137 Views

hi guys,

about TRIM, here is what i think: If the manufacturer in this case Intel delivers a TRIM program, it will work for stand alone disks, like in a portable.

In a raid config, not possible. So it will be impossible for Windows to overcome the controller to get to the cells needing to be erased, to be prepared as "virgin" again. At least not possible yet, right?

Why do companies like Perfectdisk not offer this in there programs. There should be a way that when consolidating free space (space that is deleted but not erased) would be erased, in fact why can Windows not do this?

This would be the simpliest way. Instead of deleting a record, why can't it be erased. You skip the delete and gain speed.

Saying that the SSD controller does not allow this, well then the software on the controller should be changed so this can happen.

Jeff

idata
Community Manager
137 Views

jeff_rys wrote:

about TRIM, here is what i think: If the manufacturer in this case Intel delivers a TRIM program, it will work for stand alone disks, like in a portable.

In a raid config, not possible. So it will be impossible for Windows to overcome the controller to get to the cells needing to be erased, to be prepared as "virgin" again. At least not possible yet, right?

Passing TRIM through from the RAID controller to the drives should be trivial for RAID0 and RAID1 (and RAID10). But for RAID3, 5, & 6 it is far more complicated, and would probably be incompatible with some drives depending on how well they followed the spec, and be far more sensitive to partition alignment. Also, the only way you could do RAID 3, 5, or 6 with TRIM would require reading the beginning and end of each file fragment before it is TRIMed to calculate new data for the parity stripe based on the TRIMed data being zeroed, and a reWRITE of the parity stripe which isn't necessary without TRIM, yielding performance penalties that probably outweigh the performance gains from TRIM, and extra drive wear eating into the lifespan advantage of TRIM (but probably not enough to eliminate that advantage).

So while these other RAID configurations could be tweaked to be TRIM compatible, it really only makes sense to use it in a RAID1, RAID0, or RAID1+0 config.

idata
Community Manager
137 Views

There is no vendor nor SSD drive anywhere that has TRIM working with a RAID device. Period. In order for TRIM to work with RAID, then the hardware vendors producing raid cards would have to update their firmware also. The only vendor i know that produced a SSD RAID is OCZ and you see where that product is on their spectrum products.

The s/w vendors and h/w vendors haven't gotten TRIM working on single drives reliably. Why would one think that they have gotten RAID working which is way more complicated.

And the answer is of course Intel is aware, but don't look for it anytime soon.

Reply