GA-X58A-UD5 + Marvell SATA3 Issues?

[Zarach]

Hey all,

I've just put together a new system with the following:

Gigabyte GA-X58A-UD5
Intel Core i7-930 Bloomfield 2.8GHz LGA 1366 130W Quad-Core
G.SKILL Trident 6GB (3 x 2GB) 240-Pin DDR3 SDRAM DDR3 2000 (PC3 16000) Triple Channel Kit Desktop Memory Model F3-16000CL9T-6GBTD
CORSAIR CMPSU-850HX 850W ATX12V 2.3 / EPS12V 2.91 80 PLUS SILVER Certified Modular Active PFC Power Supply
2x Western Digital 600GB 10,000RPM Raptors on the SATA3 port in RAID0
Sapphire ATI Radeon 5870

After installing Windows 7x64 I started having random issues that I have narrowed down. Windows would install fine, but would not load properly on a warm reboot. Sometimes it would give me a blank screen right before the "Microsof Windows" text appeared. Other times, I'd see "Microsoft Windows", then the system would stop responding. Note - it would not lock up. If I hit control-alt-delete it would restart. Numlock on the keyboard would respond, and if I tapped the power button it would turn off (rather than a hard lock which requires you to hold the power button or switch the power off at the p/s). On a cold boot, Windows loads quickly and runs stable for days. Only on a warm restart will these issues occur. I have talked to gskill to get the RAM timings/adjustments. Same issues. I have talked to Gigabyte, who of course blamed the RAM. Also just for the sake of argument I installed the latest BETA BIOS which is F6E. Same results. I have a hard time believing that I could have bad hardware since on a cold boot Windows runs rock solid for days. This odd behavior seems like bios/firmware/driver related nonsense?

I took the PC completely apart except for the hard drives and 1 2gb RAM chip. Same issues. I swapped out RAM, tried different slots, etc. Last night I decided to plug an older 80GB SATA 2 drive into the Gigabyte controller (not the Marvell) and I installed Windows on that. The Marvell is currently disabled in the BIOS. I have had no issues since. Obviously I'm down to the hard drives or the controller. I have seen several people having odd issues with this Marvell controller on these forums. Has anyone else run into these kinds of issues?

[brokensoldier]

This will be a short post, but I am having a similar problem.
I also have a sata 3 ssd, the Crucial Real SSD 128, and whenever I apply any overclock change to the system, even a tiny one, the computer loses the oc on reboot.  When I apply an oc, then log into windows, then power down, then power up it keeps the oc.

I changed the ram, ran memtest, ran large fft prime 95 without any errors.

After reading your experience where you swapped the ssd for a spinner drive I am inclined to believe that gigabyte boards do not react well with ssd, and that this is infact a bios issue.

I have the 890gpa-ud3h.

BOOOO!  BAD GIGABYTE!!!

[onemilimeter]

So far I think there are two problem of Gigabyte X58A board which are not solved yet, i.e. the Marvell SATA3 issue and the electrical noise issue.

[Dark Mantis]

Yes, sure there are issues with the X58A boards but that is the same with most new platforms. The unnacceptable thing is not doing enough to fix them. The Marvell chip problems beiong a good case in point. It actually runs slower especially in RAID configuration than the ICH10 SATA2 controller. They need to address these issues rapidly because this is one of the main selling points of these motherboards.

[F5BJR]

My expérience !!!

********************************************
*                                                                                    *
*    MARVELL 88SE91xx FIRMWARE STUPID      *
*                                                                                    *     
******************************************** 

WITH :

little ASROCK PCI-e card price < 20 $ ( 88SE9123 )

little ASUS U3S6 Pci-e  + USB3 ( 88SE9123 )

no RAID but my 2 enclosures with each 4 disks are all disks detected  ( total 8 disks )

IMAGE 1

http://img15.imageshack.us/i/image1kef.jpg/

***************************************************************************************************

WITH :

1 - IOISATA 88SE9128 PCI-e CARD ( RAID option )

2 – GIGABYTE GA-X58A-UD5 same port 88SE9128 same firmware same problem …

Same enclosures but ONLY 7 disks detected by WINXP

IMAGE 2

http://img208.imageshack.us/i/image2dr.jpg/

***************************************************************************************************


BUT MRU detect 8 disks !!!!

IMAGE 3

http://img31.imageshack.us/i/image3ya.jpg/

***************************************************************************************************

Why this big stupidity ?

I have also RR622 Rocket Raid RR622 no problem hardware is the same with a little I2C Philips

Firmware is make by HighPoint , and i can use in RAID 0 , 1 , 5 , 10 !!!


***************************************************************************************************


With the little ASUS or ASRock PCIe card i have also a beautiful bug :

-  if i plug in same time the RR622 card , the disks connected to the RR622 are not detected !!

- Sure that is the Marvell firmware because after a little modification of the firmware i can see my disks connected to the HighPoint card ... 

Why Great Company as GigaByte or Asus use this very basic firmware ?

***

Pierre

 

[Dark Mantis]

Well checked out Pierre. As I have said all along the Marvell firmware is crap. I just can't understand why Gigabyte(or Asus etc) put up with it.

[onemilimeter]

Well checked out Pierre. As I have said all along the Marvell firmware is crap. I just can't understand why Gigabyte(or Asus etc) put up with it.

Has Gigabyte or moderator here issued any statement about the problem? How's about Asus? As SATA3 is a main selling point, I think this issue is more critical than the "electrical noise". Frankly, one of the reason I get GA-X58A-UD3R is it's advertised to have TRUE USB3.0 and SATA3 (support RAID). I really hope that Gigabyte will solve it soon.

[absic]

This SATA 3 issue isn't just confined to the Intel platform, I did some experimenting on my AMD system  and could not get the Marvell Ports on my GA-790XTA-UD4 to run properly. In fact I have had to revert to running in Native IDE Mode as AHCI or RAID 0 just kept crashing. I tried 3 different BIOS version (including the latest beta version) and 2 different CPU's, Phenom ii X4 965 C2 and the Phenomii x6 1090T. Both were tested with BIOS at Optimized Defaults and I was trying to run Windows 7 64 bit.

As SATA3 is a main selling point, I think this issue is more critical

I have to agree with this, SATA 3 is a major selling point and, although I am realistic enough to know that sometimes new technology takes a while to settle down, it would be nice to get a solution to this and soon.

[onemilimeter]

Well checked out Pierre. As I have said all along the Marvell firmware is crap. I just can't understand why Gigabyte(or Asus etc) put up with it.

Do you think it's a problem of the firmware (software?) or the hard-wire design of the Marvel chipset? If it's the latter one (e.g. limited by the hardware design) then again it's gonna be a bad news to us...

[absic]

To be totally honest, I'm not sure what to think.

If it is just a firmware issue then it would point to a problem with the Windows 7 64 bit OS, having tested with 3 different versions of BIOS and 2 different CPU's that is the only thing, that I can think of that would have a bearing on the Marvell Chip working properly, that remained common throughout my tests. I didn't test with XP, Vista, 32 & 64 bit versions or with Linux. However, I have now spent a lot of time trying to figure out what the problem is, but I am not an expert with unlimited resources and have now run out of ideas. I need a system that is stable and reliable and the only way I can get that, at the present time, is running my rig in Native IDE Mode. This is not a big issue for me but, for others it would be.

If it is a hardware design fault, is it across the board with all the  early Marvell SATA 3 chips? The one on my AMD board is the 9128. If this chipset is the problem then I, along with many others, would want our money back, as we purchased these boards on the strength and promise of SATA 3, which is maybe why, there is no good response regarding this issue from the manufacturers. A case of  "Bury our heads in the sand and the problem will go away." I hope not but that is how it is appearing at the moment.

[Dark Mantis]

Yes I am afraid that I tend to agree with you on this issue. I am not sure about it being a hardware problem and am happy to give them the chance to prove that it is just a firmware fix that is needed, but as I have heard nothing at all on this front I ,too am begining to think that maybe it is not fixable except by replacement of the chips/boards.

It would be an idea if Gigabyte could give us their official take on this matter through runn3R. After all it does affect 100% of the latest boards purchased.

[onemilimeter]

Yes I am afraid that I tend to agree with you on this issue. I am not sure about it being a hardware problem and am happy to give them the chance to prove that it is just a firmware fix that is needed, but as I have heard nothing at all on this front I ,too am begining to think that maybe it is not fixable except by replacement of the chips/boards.

It would be an idea if Gigabyte could give us their official take on this matter through runn3R. After all it does affect 100% of the latest boards purchased.

I do not follow closely the threads which reported the issue of the Marvell SATA3 chip but it's related to my board too and thus the problem is not solved it means I will have the same problem in future. Since I do not follow closely the issue, I'm not sure if Gigabyte staffs in this forum have responded or not to this problem. I was more active in the dkslim's thread about "electrical noise" and if you go through the entire thread you will find out that the Gigabyte staff only responded after customers have made 164 posts!

Can we treat this forum as one of the official Gigabyte technical support channel? It seems that the Gigabyte technical support (especially the online http://ggts.gigabyte.com.tw/Default.aspx) is rather slow. The PSU was at first to be blamed to cause the "electrical noise" issue, what's about the Marvell SATA3? Due to the problem of HDD, SSD, or operating system?

How many Gigabyte staffs that moderate this forum? So far I only see runn3R's replies. If it's only runn3R who is in-charge of this forum, then we can imagine how busy he is. However, we really hope that runn3R will spend little time on this issue and at least tell the forumers whether they've the similar problem or not.

[absic]

Can we treat this forum as one of the official Gigabyte technical support channel?

I think you will find that this is an Official Gigabyte Forum that is watched over by them, but it is not an "Official" Tech Support Site. Runn3r is the Forum Moderator and works for Gigabyte, also there are a few other Forum members who work for Gigabyte.

But the vast majority of the members on this forum are just Gigabyte end-users, trying to solve any problems they encounter or trying to help others by offering advice from their own experiences.

[Dark Mantis]

Yes, absic115 is exactly right and even though it is designated as "official" and is watched over by runn3R (the forum moderator and the only one paid by Gigabyte)

Quite honestly I haven't seen any response from the manufacturer regarding the lack of support for what is advertised as SATA3. There is absolutely no way that I can get four times the speed of SATA2 inRAID0as the sales pitch is so fond of saying.

I do think that it's time that runn3R gave an official company statement regarding their intentions of putting this matter right. That's your que runn3R

[F5BJR]

**

i have started a little firmware modification to my ASROCK little PCI-e card ( easy because EEProm is in DIP8 )

http://img31.imageshack.us/i/asrock.jpg/

**

1 - i can mount RAID 0 , 1 , 5 , 10  exactly as  RR622 HighPoint

2 - with WINXP card is detected as a RR622

3 - after problem with WebGui and WINXP because cannot find RAID disk connected to the AsRock card

but if i connect sata cable to a RR622 the RAID mounted by the ASROCK is a normal RAID5

**

that is a just a beggining ...

**

Pierre