Is that with the latest firmware for the marvell controller?
Do you remember your scores before the update?
 

JJVAL

Heres 3 VS 6G sata ports with a 6g sata Crucial 256 gig:)
6G


3G

Here's mine.
I ran some tests with the Marvell driver, but the damn thing kept causing BSOD. 
Definitely feel that the Marvell controller is mediocre at best. It's a bit unfortunate since SATA6Gb + USB3 was the deal maker for me on purchasing this board. Might as well have bought the vanilla Classified..
 
C300 (256GB) on SATA3Gb using IASTOR drivers.

 
 
C300 (256GB) on SATA6Gb using default drivers.

 

Will Marvell ever get this resolved?
I should have researched more into issues such as this before making the purchase... So I guess it's my fault.

Does anyone have the link to the orig drivers from evga to use with the orig firmware?  The cd only has the .exe installer for the Marvell.

lerb50

I have tried to update the marvell bios (with no drives attached) with no success all it says is; file not found, syntax error, and invalid directory!
And yes i did download the cd version ( at least thats what the link said)
anyone got any help ?

I have the same situation!
I will update the BIOS but alwas the message: file not found, syntax error, and invalid directory!

m3thod

Also since i've been beating my head on this for awhile and reseraching it looks like there should be an option to change from IDE to AHCI for drives connected to the Marvell ports.  Asus seems to do this through their bios but in our bios all we have is Enable and disable for the Sata 3.0.  Is this a missing feature in the Evga Bios or in the Marvell Bios (and how do we fix it)?
edit:
Alright I finally got windows installed after removing all other hard drives.  Apparently Windows setup was getting confused with having other drives connected?  Anyways... I just ran ATTO and my writes don't get above 146.  My reads seem to be on par.  I believe this is because its currently using IDE and not AHCI?  Has anyone got their SSD to use AHCI with the new firmware?

Same issue here with the AHCI to IDE switch after update.  BIOS doesn't give you the ability to make the switch once the firmware is updated.  So I guess I wait unitl a new BIOS is released that address the ability to change the Marvell chipset from AHCI to IDE to RAID like all of my other boards...

So i just installed a Corsair P3-128 SATA3/6 SSD (Single drive) and connected it to the Marvell controller on my Classy3.
 
No option for AHCI as suspected and confirmed by others. Firmware and drivers are the lastest found here on EVGA.
 
Write-speed is close to what this drive is capable of (210mb/s) at avg. 180 mb/s.
 
Read-speed is far off though at avg 260 mb/s and this drive can do up towards 410mb/s.
 
C'Mon EVGA, pls give us some information on this. Are you working with Marvell to get AHCI-support for the SATA3/6 controller? 

It would be good to know if there are any other board manufactures that have already updated their boards SATA III controllers or if every other board manufacture is in the same boat as eVGA.
 
If it is the latter then with the release of the new Marvell 9182 chipset on the G1 Assassin i very much doubt that a firmware update will fix all these issues and it is down to a design fault just like the recent Sandy bridge issues Intel had.
 
Only difference is Marvell do not have the balls (or money) to do a mass recall.
 
I for one will be seeking a full refund from eVGA for my E770 if this does not get fixed in the next 4 weeks. 

doorules

guys, just a little reading around the internet would have told you the marvelle ports are close to worthless aa performance ports for ssd's, most especially for raid, this has been well known and documented for months

 
So based on this we can assume that there is a design issue with this controller and if so it WILL NOT be fixed with any firmware updates.
 
If this is true eVGA need to replace these boards with new boards with the new Marvell controller OR refund people.

doorules

guys, just a little reading around the internet would have told you the marvelle ports are close to worthless aa performance ports for ssd's, most especially for raid, this has been well known and documented for months

Yes, that's also why i asked and checked on this forum ;) The reply i got was that Raid is the problem, a single drive should work fine.
 
However, My starting point is that i do not believe EVGA (or Marvell for that matter) intentionally lie or misinform about their spec and/or performance, i do think (or hope might be a better word) that this will be resolved shortly. AHCI is as we all now a major feature for high performing SSD's and updates on firm- and software is needed sooner than later.
 
What concerns me is the lack of information and feedback from EVGA on this. There are many direct questions in this thread that have been left unanswered by EVGA. It gives me a feeling that they are hoping the problem goes away by itself. It is not. I have already started looking at a new mobo (not EVGA for the 1st time in many years) with documented good SATA 3/6 performance, but i will give it a little more time. Obviously the entire SATA 3/6 things seems to have caused EVGA some headache not performing as expected.

just about all the questions you guys need answered have to come from Marvelle, they make the chip in question, not EVGA, I would think EVGA just wishes they never heard of it

doorules

just about all the questions you guys need answered have to come from Marvelle, they make the chip in question, not EVGA, I would think EVGA just wishes they never heard of it

 
The Marvell controller sits on my EVGA MoBo, let's say the Marvell controller suddenly starts to burn and dies. I will not contact Marvell directly aboout this will i, no i will go to evga and request my board replaced. The anology goes for performance issues as well, i understand that the solution must come from Marvell, but i will not be the one adressing it to them, that will have to come from EVGA.
 
Obviously this controller has brought problems to EVGA, but it will be EVGA responsibility to stand behind their descision to spec the boards with it. I am a big fan of EVGA and their CS has always impressed me, i am in general a happy EVGA-customer. But that on the other hand als makes me expect them to handle this matter as well.

perptt

doorules

just about all the questions you guys need answered have to come from Marvelle, they make the chip in question, not EVGA, I would think EVGA just wishes they never heard of it

The Marvell controller sits on my EVGA MoBo, let's say the Marvell controller suddenly starts to burn and dies. I will not contact Marvell directly aboout this will i, no i will go to evga and request my board replaced. The anology goes for performance issues as well, i understand that the solution must come from Marvell, but i will not be the one adressing it to them, that will have to come from EVGA.

Obviously this controller has brought problems to EVGA, but it will be EVGA responsibility to stand behind their descision to spec the boards with it. I am a big fan of EVGA and their CS has always impressed me, i am in general a happy EVGA-customer. But that on the other hand als makes me expect them to handle this matter as well.

+1 from me too

I recived another reply from EVGA-support saying that in order for me to be able to increase HDD-speed i will need a bootable PCIe-drive!!!!
 
It seems, unless we receive dramatically new info from EVGA (which i doubt), that it's safe to conclude that the Marvell SATA 3/6 controller EVGA is using is nothing but a POS, not only for RAID but for single drives also. It is advertised as SATA 3/6, BUT DOES NOT SUPPORT AHCI, and delivers speeds as the Intel SATA2-ports.
 
So the customers that bought either the SR-2, Classified 3, FTW3 or SLI3 boards in the belive that they would receive SATA 3/6 benefits have basically been fooled. It performs on par with SATA 2 speedwise, and is missing the maintanence-features and benefits of AHCI which the Intel SATA2 has.
 
This is leaving a really, really bad taste in my mouth right now. I am hoping somebody can prove this wrong.