New SSD, Should I update chipset?

[m33pm33p]

Just finished building my new rig which includes a 60GB vertex 2. Ive read many place that you shouldnt update the chipset drivers because TRIM would not function afterwards. Is this true? Im currectly running an M4A89TD PRO/USB3 with a phenom 965. Appreciate the help.

 

[taltamir]

You know, I can't for the life of me remember if it was just when in RAID mode (but not member of array) that you lose trim, or even in AHCI mode?

 

[tweakboy]

Yes of course get latest chipset drivers of course. To me that is like installing a video card and not installing their drivers. It sure wont hurt anything, and if anything it may speed up or compatibilize your SSD to the OS. Thanks gl,

 

[poohbear]

ur ocz2 vertex has its own internal garbage collection that is just as effective as TRIM, so yes go ahead and update ur chipset drivers as u dont need TRIM.

 

[6311]

Can you give photo results with the software?

Use of this software
CrystalDiskMark
http://crystalmark.info/software/Cry...k/index-e.html

 

[taltamir]

Yes of course get latest chipset drivers of course. To me that is like installing a video card and not installing their drivers. It sure wont hurt anything, and if anything it may speed up or compatibilize your SSD to the OS. Thanks gl,

except, the non MS Drivers supposedly have problems with TRIM support. so it is nothing like installing video card drivers.

ur ocz2 vertex has its own internal garbage collection that is just as effective as TRIM, so yes go ahead and update ur chipset drivers as u dont need TRIM.

this is just flat out wrong. GC is not nearly as good as TRIM.

 

[m33pm33p]

except, the non MS Drivers supposedly have problems with TRIM support. so it is nothing like installing video card drivers.



this is just flat out wrong. GC is not nearly as good as TRIM.

You sound like you know what Im asking, as what you just said sounds exactly like what I thought (If I just made any sense?)

 

[taltamir]

Update is a little misleading, there is an actual MS written "generic driver" for SATA controllers (well, AHCI actually), if you install the driver from the mobo maker, it will replace that with a driver written by intel, AMD, nvidia, SIS, etc. Those drivers might or might not be newer, but even if newer, they might not be actually better.

The MS drivers were first to support TRIM, followed by intel. the MS driver is also very robust.
This thread discusses it in detail:
http://forums.amd.com/forum/messageview.cfm?catid=383&threadid=126518
with the answer is given by gaspard.leon

Basically there are two issues at stake here:
1. Does the driver support TRIM, at all?
2. How does the driver handle TRIM when in RAID mode?

To answer #1:
MS and intel drivers support TRIM, AFAIK they are they only ones. Windows, Linux, and Solaris support TRIM, MacOS does not.

To answer #2:
No controller supports trim on an SSD that is part of a RAID array, supposedly it is possible, but requires modifications to the controllers (and are sure to cost extra), as of yet nobody offers a controller that can do so.

But there is another issue, Say you have set your controller to "RAID Mode" and have made a RAID1 array of 2x2TB HDDs, and you have your OS on an SSD NOT part of a RAID array. Well, because the CONTROLLER is set to RAID mode that would cause TRIM not to work... with the exception being intel mobo controller that is using the latest intel drivers (it wasn't available at first, but was patched in)

So, what does all this mean to you? In theory you should have working trim by installing windows 7 and not touching the chipset drivers from AMD. And unless they have recently released new drivers that finally support it, installing the AMD chipset drivers will disable your TRIM, so you should NOT install them.

 

[m33pm33p]

Update is a little misleading, there is an actual MS written "generic driver" for SATA controllers, if you install the driver from the mobo maker, it will replace that with a driver written by intel, AMD, nvidia, SIS, etc. Those drivers might or might not be newer, but even if newer, they might not be actually better.

The MS drivers were first to support TRIM, followed by intel. the MS driver is also very robust.
This thread discusses it in detail, basically the answer is given by gaspard.leon

Basically there are two issues at stake here:
1. Does the driver support TRIM, at all?
2. How does the driver handle TRIM when in RAID mode?

To answer #1:
MS and intel drivers support TRIM, AFAIK they are they only ones. Windows, Linux, and Solaris support TRIM, MacOS does not.

To answer #2:
No controller supports trim on an SSD that is part of a RAID array, supposedly it is possible, but requires modifications to the controllers (and are sure to cost extra), as of yet nobody offers a controller that can do so.

But there is another issue, Say you have set your controller to "RAID Mode" and have made a RAID1 array of 2x2TB HDDs, and you have your OS on an SSD NOT part of a RAID array. Well, because the CONTROLLER is set to RAID mode that would cause TRIM not to work... with the exception being intel mobo controller that is using the latest intel drivers (it wasn't available at first, but was patched in)

So, what does all this mean to you? In theory you should have working trim by installing windows 7 and not touching the chipset drivers from AMD. And unless they have recently released new drivers that finally support it, installing the AMD chipset drivers will disable your TRIM, so you should NOT install them.

Think that about sums it up. Thanks,


Edit: woot! 100 posts.

 

[poohbear]

except, the non MS Drivers supposedly have problems with TRIM support. so it is nothing like installing video card drivers.



this is just flat out wrong. GC is not nearly as good as TRIM.

no its not "flat out wrong", learn a thing or two before telling ppl "they're wrong".

Sandforce has the best internal garbage collection on the market right now, Anandtech, SSD review and other sites have demonstrated that it IS just as effective as TRIM and so users are fine ditching the crappy MS drivers. Other SSD controllers vary in result, Crucial's c300 isnt as effective, but the sandforce ones are. Please go read up on Sandforce drive's internal garbage collection before commenting others are "flat out wrong" and making blanket statements about all 3rd party garbage collection.

 

[taltamir]

no its not "flat out wrong", learn a thing or two before telling ppl "they're wrong".

Sandforce has the best internal garbage collection on the market right now, Anandtech, SSD review and other sites have demonstrated that it IS just as effective as TRIM and so users are fine ditching the crappy MS drivers. Other SSD controllers vary in result, Crucial's c300 isnt as effective, but the sandforce ones are. Please go read up on Sandforce drive's internal garbage collection before commenting others are "flat out wrong" and making blanket statements about all 3rd party garbage collection.

it IS flat out wrong. Anandtech and others have not "demonstrated it to be as good as trim". And MS Drivers are in fact, superior to all but intel's.
Garbage collection is better than nothing, but its not equal to trim.

http://www.anandtech.com/show/2829/14

Presumably this isn’t without some impact to battery life in a notebook. Furthermore, it’s impossible to tell what impact this has on the lifespan of the drive. If a drive is simply reorganizing data on the fly into a better (higher performing) state, that’s a lot of reads and writes when you’re doing nothing at all. And unfortunately, there’s no way to switch it off.

% of new Perf for Trimmed drive based on data in that article: 98.5%
% of new perf for GC drive from that article: 91%
% of new perf for used drive with neither: 70.6%

Idle garbage collection doesn't happen instantly, it takes longer then trim, doesn't provide as big a benefit, presumably consumes extra power and lifespan, and requires compatible FS, such as NTFS.

Now, GC is far from useless, its actually very good, but its not TRIM.
Finally, neither GC NOR TRIM is absolutely necessary, since without either you still have a lower limit on how low performance will drop. At its lowest, it still 70% the max speed of the SSD in question.

The above were all for indilinx...

here is for the C300, which as you mentioned is not "as effective" (an understatement):
http://www.anandtech.com/show/3812/the-ssd-diaries-crucials-realssd-c300/8

The idle GC just isn’t very aggressive. Compared to what Indilinx and SandForce have done, Crucial’s implementation just doesn’t cut it.

Sanforce you said was the best... there are two anandtech articles about it.
This one doesn't mention GC at all, and shows that with trim it immediately bounces back completely after the harshest of tests devised specifically to defeat its deduplication:
http://www.anandtech.com/show/3681/oczs-vertex-2-special-sauce-sf1200-reviewed/4

This one plain doesn't mention TRIM OR GC:
http://www.anandtech.com/show/3667/oczs-agility-2-reviewed-the-first-sf1200-with-mp-firmware/10