SSD freeze/stutter problem

R

rodneycubed

Limp Gawd
Joined
Nov 30, 2007
Messages
184
I recently purchased an intel 40gb SSD and have been satisfied with how "speedy" it is. My problem is that every once in a while it will seem to freeze (My OS) for about half a second. If I have google chrome open it will crash it too. :/

I am running windows 7 and am not too sure if anyone else has had this problem and if so, how did they remedy it?
 
Did you create a smaller partition (like 32GB) and reserve some free space? Or did you create one partition covering the entire SSD? How full is the drive now?

You may have exhausted the number of free flash cells; and cause triple wear (means only the third lifespan) and low performance (high write latency).

To rectify, use HDDErase to clear the HPA areas of the SSD (destroys all data on the SSD!) then partition leaving some space free (32GB on 40GB volume). That would help keep the performance degradation to a minimum and also have proper lifespan.

With exhausted sectors, your SSD wears at least three times as much, possibly even more. That means for each KB you write, it is like you have written three times the amount. This also reduces the lifespan of the unit to 33%.
 
Purchased it about a month ago, so hopefully I didn't degrade it too much. To answer your questions:

No, I partitioned the whole drive and I have 11GB of free space left.

What is the lifespan of an ssd?
 
lifespan is how long your product lasts. Your usage makes the SSD last much shorter; it will have run out its write cycles much sooner. Writing 1KiB to the drive may cause 128KiB+128KiB = 256KiB writes.

If you want to resolve this, you would have to use the HDDErase utility as i described and repartition (and re-install). If you don' t want to do this just leave as is and accept the lower lifetime and performance.

This is about an Intel X25-V 40GB?
 
Correct. I'm not trying to be lazy. I was just curious as to the lifetime of your typical ssd. I reformat often (relatively speaking) anyways, so I will definitely do what you instructed when I do reformat.
 
Reformatting alone will not help, you need to wipe the SSD using HDDErase, HPA Block Erase. Only that will clear the SSD.

Windows 7 is said to zero-write when formatting, though i would feel more comfortable using HDDErase to make that happen. Then create a partition smaller than the full capacity, forcing a portion of the drive to be unused. In fact, these 'unused' flash cells are used to accelerate random writes; this will both add to performance and prolong the lifetime of the SSD. Without this, the lifetime would be crippled and every write could take 256KiB of write-cycles.
 
The X25-V 40GB supports TRIM. If you're using the default MS storage controller drivers, it should be working on your drive. You can check with CrystalDiskInfo to be sure. If TRIM is working, all of the previous comments about the drive filling up aren't applicable, since the SSD's controller will clean up invalid data from pages as they're deleted.

Also, the Intel drives have about 7% free space completely inaccessible to the user - even if you fill it to the maximum 40,000,000,000 bytes, it will still have 2,949,672,960 bytes available to the drive for working space. Testing has shown steady improvement up to about 30% reserved space. It's a tradeoff between performance/longevity and how much of your expensive drive you actually want to use.

As for lifespan, Intel has said that the X25-M can handle 100GB of writes per day, every day, for 5 years. In their internal testing, most users were writing 2-3GB/day and their heaviest user was under 20GB/day. In another thread, I checked my drive 79 days after it was delivered - 1735 hours powered on and 381.53GB of writes. That comes out to 5.28GB/day for ~72 days. So if you can write 10x what Intel's heaviest user did or 40x what I did, every day, you might cut your SSD's lifespan down to 3 years. For reference, 1TB drives were just getting ready to be launched 3 years ago. Not quite 5 years ago, Seagate was launching the 7200.8 line with a 400GB model for $330. In other words, your SSD will probably be worth less than $50 when it dies, if you manage to double Intel's 100GB daily rating.

Win7 should do it automatically, but have you checked to make sure the partition is properly aligned? http://www.techpowerup.com/articles/other/157 is an easy calculator that tells you how to find the needed values. If your partition isn't aligned with the SSD's blocks, each chunk of data will actually be split over two physical blocks, increasing the amount of work being done.
 
I checked my drive with crystaldiskinfo and it says that TRIM is working.

474 Power on hours with 215GB Host writes.
 
Intall the Intel SSD Toolbox. CrystalDiskInfo just tells you if TRIM is supported on your drive not if its actually working.

Also for TRIM to work you need to have the latest Intel Firmware, which probably came shipped on the 40gb, but i would check to make sure.
 
sub.mesa said:
Reformatting alone will not help, you need to wipe the SSD using HDDErase, HPA Block Erase. Only that will clear the SSD.

Windows 7 is said to zero-write when formatting, though i would feel more comfortable using HDDErase to make that happen. Then create a partition smaller than the full capacity, forcing a portion of the drive to be unused. In fact, these 'unused' flash cells are used to accelerate random writes; this will both add to performance and prolong the lifetime of the SSD. Without this, the lifetime would be crippled and every write could take 256KiB of write-cycles.
Click to expand...

Don't the Intel drives already use free space for that purpose regardless of whether it's formatted/allocated or not? That's the impression that Anand's articles always gave me... And why he recommended leaving 20% space empty on the drives...

Anand said:
In previous articles we had to test SSDs in two conditions: new and used. The new state is just after a secure erase, the used state required us to write data to every user accessible portion of the drive first before benchmarking. The former resulted in great performance, the latter meant the SSD had to do a lot of juggling of existing data whenever it went to write. The second scenario no longer exists with TRIM. The act of formatting your drive or deleting files (and emptying the recycle bin) will TRIM invalid data.

Performance in a TRIM enabled system is now determined not by the number of invalid blocks on your SSD, but rather the amount of free space you have. I went into a deep explanation of the relationship between free space and the performance of some SSDs here.

TRIM will make sure that you don’t have to worry about your drive filling up with invalid data, but it doesn’t skirt the bigger issue: dynamic controllers see their performance improve with more free space.

My rule of thumb is to keep at least 20% free space on your drive, you can get by with less but performance tends to suffer. It doesn’t degrade by the same amount for all drives either. Some controllers are more opportunistic with free space (e.g. Intel), while others don’t seem to rely as much on free space for improved performance. Addressing performance degradation as drives fill up (with valid data) will be one of the next major advancements in SSD technology.
Click to expand...

That was taken from this article... The other article linked within that very same quote goes into even more detail about this. Edit: To add another quote from that other (older) article...

Intel’s controller is dynamic, it uses the entire drive as spare area until you’ve written every LBA once. Then it relies on the remaining 7.5% of non-user-space as its scratch pad. That’s why its new, out of box, performance is so good.Other controllers may not be quite as dynamic, but they may also take a smaller performance hit when fully used. Why would Intel work so hard to make its out of box performance so high, even when it’ll be short lived? Because of TRIM.

<skip ahead to his introduction of TRIM aaand...>

But remember, today’s consumer drives only ship with roughly 6 - 7% spare area on them. That’s under the 10% minimum even from our more sophisticated controller example. By comparison, the enterprise SSDs like Intel’s X25-E ship with more spare area - in this case 20%

What TRIM does is help give well architected controllers like that in the X25-M more spare area. Space you’re not using on the drive, space that has been TRIMed, can now be used in the pool of replacement blocks. And as IBM’s study shows, that can go a long way to improving performance depending on your workload.
Click to expand...

So the way I see it, having completely unallocated space as scratch area would only be beneficial if you're not running TRIM (if running a RAID for instance) or if you're using a SSD other than the X25-M which may not use free space as aggressively as the X25-M.
 
Last edited:
nitrobass24 said:
Intall the Intel SSD Toolbox. CrystalDiskInfo just tells you if TRIM is supported on your drive not if its actually working.

Also for TRIM to work you need to have the latest Intel Firmware, which probably came shipped on the 40gb, but i would check to make sure.
Click to expand...

Aren't you confusing the DisableDeleteNotify console command w/the way CrystalDiskInfo actually reads whether TRIM is enabled or not? I'm quoting the guy that figured out how to enable the X25-V firmware on the 40GB Kingston drives for TRIM here:

Telnet said:
DisableDeleteNotify just indicates that TRIM is enabled in the OS, not necessarily in the drive. The real test is reading the value of Word 169 from the drive's IDENTIFY DEVICE data that really tells if TRIM is enabled or not. CrystalDiskInfo reads this value,
Click to expand...

Source.
 
Impulse said:
Aren't you confusing the DisableDeleteNotify console command w/the way CrystalDiskInfo actually reads whether TRIM is enabled or not? I'm quoting the guy that figured out how to enable the X25-V firmware on the 40GB Kingston drives for TRIM here:



Source.
Click to expand...

Try it on XP (which doesnt have TRIM)

Crystal disk just checks the drive for features available.
 
Impulse said:
Don't the Intel drives already use free space for that purpose regardless of whether it's formatted/allocated or not? That's the impression that Anand's articles always gave me... And why he recommended leaving 20% space empty on the drives...

So the way I see it, having completely unallocated space as scratch area would only be beneficial if you're not running TRIM (if running a RAID for instance) or if you're using a SSD other than the X25-M which may not use free space as aggressively as the X25-M.
Click to expand...

From your own quote...
Intel’s controller is dynamic, it uses the entire drive as spare area until you’ve written every LBA once.
Click to expand...

Any free "user" space is used as spare area only until it's been written to. If you have a 40GB partition but only 30GB of data, you've got that extra 10GB to use as spare area. However, if you manage to temporarily add another 10GB of data there (using your system as a big "flash drive" to move 10GB of backups or something), then it will no longer be used as spare area, even after you delete the files and free the space back up. Filling the drive up once will permanently (until you erase the whole thing again) reduce you to the default ~7% spare area. Leaving some unpartitioned space will ensure that you never remove that space from the spare area pool.

But then again, TRIM does vastly reduce the need for spare area, since any deleted files should have their corresponding blocks erased quickly, giving you all those pages available for immediate writing. If you have enough regular blocks available for writing, then you don't need to use spare area.
 
Click the links thru and follow that quote to it's source, you took it out of context. He was describing that behavior in the article PREVIOUS to introducing the concept of TRIM, read some of the quotes that follows it... He states that once TRIM is in place the controller will use any free space on the drive as scratch space thanks to TRIM. 'Least that's the way I read it, could be wrong...

Space you&#8217;re not using on the drive, space that has been TRIMed, can now be used in the pool of replacement blocks. And as IBM&#8217;s study shows, that can go a long way to improving performance depending on your workload.
Click to expand...

It'd be kind of stupid if the controller itself couldn't recognize TRIM'd space and use it dynamically...
 
It makes sense to me. I remembered reading the original quote before, and totally missed that updated note.
 
fullvietFX said:
I just reformatted and the problem still persists.
Click to expand...

Did you just reformat, or use SecureErase(found in OCZ Toolbox) to reset the drive? Are you in AHCI, or IDE mode? If in AHCI, which drivers are you using? Do you have any kind of power saving features on?
 
SicKlown42012 said:
Did you just reformat, or use SecureErase(found in OCZ Toolbox) to reset the drive? Are you in AHCI, or IDE mode? If in AHCI, which drivers are you using? Do you have any kind of power saving features on?
Click to expand...

Reformat with HDDErase 4.0. Currently using AHCI with Intel RST. Power saving is on Balanced in Windows.
 
Did you restore an image for the OS or install from scratch?

How full is the drive?
 
Go to the Power settings and select Change Plan Settings on the Balanced plan. Then select Change Advance Settings and make sure the HDD section is not allowed to sleep. Apparently that can cause some performance issues with Sandforce SSDs.

Also, are you sure your drive is aligned? AS-SSD will show you a number the top left corner, which should be a number dividable by 4096 and in green writing.

You could also check in Device Manager and make sure the SSD has "Enable write caching on the device" option in the Policies tab selected.
 
You must log in or register to reply here.
Back
Top