Jump to content
Corsair Community

Force 3 120gb 5.05a firmware problem


ange30

Recommended Posts

I just updated the 5.05a firmware today all went nice and smooth,what confuses me is after updating my Corsair Force 3 120gb ssd the updater tool shows firmware 5.05 not 5.05a same goes with Intel rst ,assd they all list my drive with firmware version 5.05 not 5.05a,i just want to ask if there is a possibility that the wrong firmware was uploaded?
Link to comment
Share on other sites

  • Administrators

Please attach a screenshot of the FW Rev. in your update tool.

 

Also Four X, didn't you say in the other thread that you couldn't update your firmware because of the signature? Do you have more than one Corsair SandForce based SSD?

Link to comment
Share on other sites

Please attach a screenshot of the FW Rev. in your update tool.

 

Also Four X, didn't you say in the other thread that you couldn't update your firmware because of the signature? Do you have more than one Corsair SandForce based SSD?

 

yep i do have two of them, one in my laptop an Force 3 and one in my Desktop PC an Force GT.

 

here's a pic showing my FW Rev.

 

http://i43.tinypic.com/jj4nyf.png

 

i'm not having a very good FW day guys. haha :biggrin:

Link to comment
Share on other sites

Thanks a lot RAM GUY for looking into this problem,maybe the wrong firmware was uploaded,just guessing here not that i have any issues with 5.05 firmware everything is working fine, but please do post an update on this matter as soon as you can.

Many thanks!!!!!!!!

Link to comment
Share on other sites

Same here, I have 2 Corsair SSD, this one is a Force GT and the other one is a Force 3.

 

Only posting one image as both drives display the same issue.

 

http://i.imgur.com/HoMVKkk.png

 

 

Yeah, I deleted the S/N, I'm paranoid:wall:. Thank you Four_X for your S/N btw:naughty:.

Link to comment
Share on other sites

  • Corsair Employee

This was the reply I received from our Engineering so as I was about to post before it will be the same F/W. just depends which specific Drive you have as to it reporting F/W 5.05 orr 5.05a.

Actually, the FW 5.05 is the latest version for all Nova2, Accelerator, Force 3 and Force GT (25nm Intel/Micron). FW 5.05A is for Force GS (24nm) only. Both 5.05 and 5.05A are identical in functionality but the different is the NAND support.
Link to comment
Share on other sites

This was the reply I received from our Engineering so as I was about to post before it will be the same F/W. just depends which specific Drive you have as to it reporting F/W 5.05 orr 5.05a.

 

Thanks for clearing that up for us RAM GUY :biggrin:

Link to comment
Share on other sites

  • 3 weeks later...
After updating from 5.03 to (supposedly) 5.05a, the SSD's firmware version shows as 5.05 (no a), is this normal?

SSD is Corsair Force GS 360 GB (SATA3) used on system in my profile.

Wanting to combine two such SSDs in RAID0, but one shows as 5.05 (older one) and one as 5.05a (purchased this month), so not sure I should proceed.

 

Actually, the FW 5.05 is the latest version for all Nova2, Accelerator, Force 3 and Force GT (25nm Intel/Micron). FW 5.05A is for Force GS (24nm) only. Both 5.05 and 5.05A are identical in functionality but the different is the NAND support.

 

The interesting point is that you have the Force GS, what should SHOW 5.05a, not 5.05 and this tells me i was right when saying:

 

I guess the field showed in the update is just float int value

 

The tool simply cant show the a in 5.05a because its formated on the old-drive in only numbers allowed (no character-string), while the new drive with 5.05a from production may save the value as string.

 

Or simply the updater always writes new value to only 5.05 whatever it is 5.05a.

 

The RAID should work, since high likely both have 5.05a, because authentic 5.05 (non a-revision) would not support ForceGS chips / whatever 5.05 and 5.05a hold same functionality for different chips -> ensure you have same chips/drive.

 

PS: I really like to know why you people have that particular question in mind at all when everything runs fine!? there is noone can guarantee RAID can work, in doubt you have to ask the Sata-Controller RAID-driver vendor if that particulary drives with that particular firmware in an particular raid mode (likely 0) would work... and that means no guarantee, try and if it wont work review drivers changelog, on the other hand when a disk-manufacturer states "RAID-ready" on a drive (may be only for particular controlers), he has to ensure that the drives work, so RMA would be valid, whats not the case for this drive. but not to frighten you, i see no point why it not should work, it "SHOULD" work. you already know that, thats why you bought 2 of the same drive O.o.

 

my suggestion is not to use a raid on SSDs. why? because raid were cool when you had slow HDD and there was no other way of increasing speed. nowadays you can simply buy a greater faster drive or buy PCIe-Cards that have predefined internal raid0 if you really need to operate at such ultra-performance levels. sry that i have to mention the sams. 840pro, its so amazing. still, the jump from HDD to SSD was significant in dimension 10x. the only way i could say i noticeable see difference between a single SSD and a 4x raid0 would be a benchmark result value O.o....

 

the real problem in RAID is 99% of the times (when you not pay lots of money) it is a SOFTWARE raid, what means, it has no own hardware-chip doing the math, yes DESPITE/ESPECIALLY WHEN ITS A MOBO-CONTROLLER, that just holds the instruction set, causing your CPU to extra-process any read/write that, simply decreases CPU ICs caches (L1,L2,L3) bandwith, so the marginaly read/write improvements that you might not notice come at a high cost that you might notice! this issue weight heavier the faster the raid-drives are. keep that in mind.

 

for superior performance on certain apps or games i suggest you lots of ram and a permanent RAM-disk.

 

if you need more speed on your storage-files, you can still buy RAID-prooved-HDDs.

 

Im sry i dont know of any SSD with RAID labeled.

Link to comment
Share on other sites

Software vs hardware RAID is mostly irrelevant when you are talking about simple RAID-0 (what people are usually doing with SSDs) & RAID-1 (likely done on a server) as there are no parity calculations involved. There should be no significant overhead to limit performance on these simple RAID's, and performance issues are generally due to firmware design of the drives (which updating might improve) or the drivers for the RAID controller.

 

Intel RST RAID in particular is very finicky about drive combinations; to illustrate this simply change an Intel RST RAID set showing performance issues into a dynamic disk pair and see if the issue dissappears (not an option for boot drives). If it does, the issue is more significantly the RST drivers vs the drive firmware. This performance throttling on RST is very common with RAID-1 (usually most dramatic on sequential read vs write in Crystal Diskmark), particularly with Seagate magnetic drives (regardless of whether they are enterprise or consumer lines).

 

When more intensive RAID involving parity calculations such as RAID-5 is involved, it is relevant as the hardware RAID controller would offload those parity calculations from the processor vs a software RAID cfg.

Link to comment
Share on other sites

My acid test of a RAID-0 (or RAID-1) is to run PCMark's hard drive / storage tests on the array vs a single drive. If the score decreases in any RAID0/1 cfg vs the score of a single drive from the array, you should use a different method as some element of the RAID set is not a good match. Even in RAID-1, any competent controller (including Intel RST) will distribute QD32 type multithreaded reads across both drives in the array resulting in higher PCMark scores. If it doesn't, use a different configuration or else you're kidding yourself if you think you've created a faster setup in real world use. Pretty sequential numbers are the least relevant element of a storage subsystem, unless you sit around copying videos and ISOs all day.
Link to comment
Share on other sites

  • Administrators
The interesting point is that you have the Force GS, what should SHOW 5.05a, not 5.05 and this tells me i was right when saying:
I guess the field showed in the update is just float int value

The tool simply cant show the a in 5.05a because its formated on the old-drive in only numbers allowed (no character-string), while the new drive with 5.05a from production may save the value as string.

 

Or simply the updater always writes new value to only 5.05 whatever it is 5.05a.

Actually, I think it's a text field, and I believe I've seen a screenshot or two showing 5.05a.

 

The RAID should work, since high likely both have 5.05a, because authentic 5.05 (non a-revision) would not support ForceGS chips / whatever 5.05 and 5.05a hold same functionality for different chips -> ensure you have same chips/drive.
Like Ram Guy said, they're identical in functionality but the difference is a tweak for the different NAND.
Link to comment
Share on other sites

Software vs hardware RAID is mostly irrelevant when you are talking about simple RAID-0 (what people are usually doing with SSDs) & RAID-1 (likely done on a server) as there are no parity calculations involved. There should be no significant overhead

 

On a raid 5 you wont need any parity calculations too. you need to calculate parity, when a disk fails and you have to replace a disk. In that case the Hardware-Controller or the CPU rebuilds the bits of the missing Disk by comparing the checksum per block with the parity-bit.

 

What you say, still, is totally wrong. Writing in a FW-Raid0 with the CPU at say 500mb/s takes a part 500mb/s bandwith of the CPU-caches/memory-subsystem-bandwith OVERALL, that means L1,L2,L3, Memory can operate at a bandwith that is 500mb/s slower than before...why, because it's busy a little with read/writing the SSDs... whatever no "calculation" is involed. Simple as that. If you still know what PIO-Mode is and what advancement DMA-mode gave... you would not like to call every write/read with the CPU... please learn some on the topic!

 

There is no calculation, there is a processor calling a read/write-instruction. The data goes from the disk into the memory-subset up into the L1-cache and back-way down to write it to memory. That is not neccesary without a FW-RAID! Because without a FW-RAID the SATA-Controller can directly write data into the memory.

 

Gladly the CPU/Memory-Subsystem of modern CPU is pretty much sufficent (see your Aida64 CPU/Mem-Subsystem Benchmark) to be capable of even writing at many gigabyte per second. Still your Application/Game may require ALL of available the bandwith to operate satisfacting!

 

Lets assume you are right.... noone could sell a Hardware-Raid-Controller... these would be overexpensed crap... but they exists still and are pretty much expensive ... and professionals do buy them... why should they do that, when for the same price the could get a high-end Intel-Board... so what? ;)

 

See you next time when solving the miracle why you encounter micro-stutter without a SLI ... might be the raid0. :D

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...