EVGA

Helpful ReplyI asked EVGA and Highpoint collaborate to fix this bizarre incompatibility.

Author
jmkuchta
New Member
  • Total Posts : 11
  • Reward points : 0
  • Joined: 2008/06/20 11:17:46
  • Status: offline
  • Ribbons : 0
2010/03/10 02:44:09 (permalink)
to: support@evga.com, support@highpoint-tech.com
date: Wed, Mar 10, 2010 at 2:33 AM
subject: EVGA X58 SLI LE incompatible with RocketRAID 3520

Hello EVGA and HighPoint,

I have used EVGA products for about three years, and Highpoint products for almost two years.  I recently decided to build a new computer and purchased an EVGA X58 SLI LE motherboard to use with my Highpoint RocketRAID 3520.  Up until that point I was using an MSI P7N Diamond motherboard, and had a few issues that Highpoint never resolved, but overall, it wasn't a big deal.  I also had one issue with an EVGA video card, and EVGA's excellent support team replaced the card quickly.

The EVGA board POSTs with the RocketRAID card from a cold boot without issue.  I am able to see all my RAID arrays, create new ones, and even install Windows onto them.  However, when I restart the computer, the motherboard doesn't post.  The mobo's diagnostic sensor reads F7 and doesn't move.  I contacted EVGA about this last week, and after trying different PCI-Express slots and changing a couple BIOS settings, they told me that I was pretty much SOL--this mobo was simply incompatible with the RocketRAID card.  EVGA also denied my request for a return or exchange of the motherboard with one that does work with the RocketRAID 3520, saying that RAID cards are notoriously incompatible and that this is the "state of the industry."

I then tried to RMA the mobo to Newegg, but the 'Egg says the mobo is non-refundable.

So, considering that the EVGA mobo is pretty sweet of its own accord and that I paid $240 or so for it, and considering that my RocketRAID card has served me pretty well and that I paid $430 for that, I would really appreciate if your engineers would put on your thinking caps and get to work on some sort of solution.  I would gladly update my mobo BIOS or RocketRAID firmware if there were something that could fix this.

I really appreciate your time, and I would love to be able to recommend both HighPoint and EVGA products to my enthusiast friends in the future.  Thank you for making quality products and having US-based tech support.

Regards,
John Kuchta
Media Engineer


NB: I have the latest firmware for the 3520 and the latest BIOS (44) for the mobo.
#1
EnochRoot
Superclocked Member
  • Total Posts : 207
  • Reward points : 0
  • Joined: 2009/08/04 15:05:30
  • Status: offline
  • Ribbons : 2
Re:I asked EVGA and Highpoint collaborate to fix this bizarre incompatibility. 2010/03/10 22:50:09 (permalink)
This is a lesson I learned many moons ago with Promise and HighPoint.  To be honest, these cards aren't touched in the enterprise because they suck even on the high end.  For RAID, LSI/3ware, Areca, or Adaptec are used.  From what I've read on linux forums, that series card is using a marvell chipset with scsi emulation in software. 

I'm sorry about your situation.  There is no solution besides a firmware update on highpoints side.  It isn't the motherboards responsibility to interpret every card according to how they felt like mapping the memory footprint.  And that's where it sounds like it's hanging.  The card is likely stomping all over the low level memory.  Which sounds like them, frankly.  They need to make it compatible with the motherboard.  Good luck, and sorry again.

--==Hardware Profile==--
EVGA Classified E760
Intel i7 960
PC Power and Cooling Silencer 910 watt
Intel Pro PT 1000 Server NIC
LSI Logic/Avago SAS2008 Fusion-MPT SAS-2 RAID
4 X Crucial M4 SSD
EVGA GTX 550
12 GB G-Skill DDR3 Tri-Channel
Renesas uPD720201 USB 3.0
OS- LINUX 64-BIT(Kernel 4.5)
#2
YerBuddy
Omnipotent Enthusiast
  • Total Posts : 8919
  • Reward points : 0
  • Joined: 2005/08/25 09:55:57
  • Location: Classified
  • Status: offline
  • Ribbons : 59
Re:I asked EVGA and Highpoint collaborate to fix this bizarre incompatibility. 2010/03/11 01:34:52 (permalink)
Good answer Enoch!
#3
jmkuchta
New Member
  • Total Posts : 11
  • Reward points : 0
  • Joined: 2008/06/20 11:17:46
  • Status: offline
  • Ribbons : 0
Re:I asked EVGA and Highpoint collaborate to fix this bizarre incompatibility. 2011/04/29 03:12:48 (permalink)
I do concede that Highpoint cards are generally not desirable. The further I can run from Marvell controllers the better. And you're probably right about what the Highpoint card is doing. In the future I'd rather buy an Areca or LSI.
 
However, that particular mobo (X58 SLI LE) has four x16 PCIe slots, and the X58 chipset doesn't officially support that, so EVGA is doing some auto-switching voodoo to allocate the PCIe lanes. I'm convinced this is more of a problem than Highpoint's firmware, because the Highpoint card works in an old Asus P5W DH Deluxe without issue, and also worked in a more recent MSI P7N Diamond mobo with more-standard PCIe lane allocations. I also got some information from a system engineer for BOXX Technologies who explained that this lane-allocating auto-switching is different on each motherboard, and some companies do it better than others. EVGA generally says their boards don't support any storage controller cards because they know the poorly-documenting auto-switching interferes with these cards' operation. However, I bet that my replacement board, an EVGA X58 SLI 3, which only has 3 x16 PCIe slots, would support the Highpoint card without issue. After all, it supports an OCZ RevoDrive X2 which is just based on the Silicon Image SiI 3124 SoftRAID 5 Controller.
 
So it's really hard for me to believe that Highpoint could actually write a firmware that would make it work in the SLI LE--because the SLI LE is doing unpredictable lane-switching and EVGA can only change that with a new BIOS.
 
I just wish EVGA and Highpoint could have told me this instead of pointing the finger and not actually helping--as nice as their respective support staffs were.
#4
_NickM
CLASSIFIED Member
  • Total Posts : 2015
  • Reward points : 0
  • Joined: 2007/05/11 15:28:19
  • Status: offline
  • Ribbons : 3
Re:I asked EVGA and Highpoint collaborate to fix this bizarre incompatibility. 2011/04/29 03:22:47 (permalink)
There is no lane switching on either the X58 SLI LE or the SLI 3. Both boards' pci-e lanes are fixed. The SLI LE has 16x8x4x8 from top to down with the x4 coming from the south bridge. For the SLI 3, the lanes are fixed at 16x8x8. This cannot be altered with the BIOS. These kinds of devices can cause issues when it comes to loading them into the Option ROM. We do try to test and support as many devices as we can but unfortunately some devices can still present issues.

 
EU Questions? Contact me @ 
nickm@evga.com | +49 89 18 90 49 - 27 
#5
jmkuchta
New Member
  • Total Posts : 11
  • Reward points : 0
  • Joined: 2008/06/20 11:17:46
  • Status: offline
  • Ribbons : 0
Re:I asked EVGA and Highpoint collaborate to fix this bizarre incompatibility. 2011/04/29 03:48:00 (permalink)
Wow, jesus, I wish someone had told me that when I filed the original ticket over a year ago. Thanks for the hard info, especially about the x4 coming from the southbridge and my new SLI3.
 
If it's an option ROM issue, does that explain why my MSI and Asus boards support it? What is different about EVGA's option ROM compared to Asus's, for example? Is it a matter of size, wherein it is more expensive to have a larger option ROM? I have the impression that TYAN and Supermicro boards have ENORMOUS option ROMs.
 
Thanks for jacking our domes with knowledge!
#6
KMoore4318
Pistoj Doulos Unum Pluribus
  • Total Posts : 17850
  • Reward points : 0
  • Joined: 2008/12/04 04:20:57
  • Location: Atlanta, Ga
  • Status: offline
  • Ribbons : 238
Re:I asked EVGA and Highpoint collaborate to fix this bizarre incompatibility. 2011/04/29 06:05:33 (permalink) ☄ Helpful

 
You do know BIOS 82 has new option rom information, in fach 80 and 81 did too, but 82 had some minor fixes, I don't use the card your talking about, so I have no idea if it's helped, but it's something new to play with, also the only boards that EVGA has that utilize the NF200 chip for multaple 16X lanes, and more PCIe that chipset possable, are the discontinued 759 with 1 chip, and the 762 with two, for 3(16X) support and 4(16X) support respectavly.   
post edited by KMoore4318 - 2011/04/29 06:08:12

  
Just clink Link to Register >  My Affiliate Code:VTJPDC4OMB 
1)965 V-8  E761(77)
2)980X AX1200 E760(77) 980 ti
3) E770 (77)2(590) SLI AX1200 
4) X299 FTW K, 9980XE 2 2080 supers on AX1200
#7
Zaskar
Superclocked Member
  • Total Posts : 126
  • Reward points : 0
  • Joined: 2007/10/14 08:31:10
  • Status: offline
  • Ribbons : 2
Re:I asked EVGA and Highpoint collaborate to fix this bizarre incompatibility. 2011/04/29 07:21:13 (permalink)
jmkuchta
If it's an option ROM issue, does that explain why my MSI and Asus boards support it? What is different about EVGA's option ROM compared to Asus's, for example? Is it a matter of size, wherein it is more expensive to have a larger option ROM? I have the impression that TYAN and Supermicro boards have ENORMOUS option ROMs.

Good question, its one of the only board brands that LSI raid cards dont really work with either, you get post issues relating to the option rom loading.
 
I guess it could be an LSI issue, but why does it work on every other board?

EVGA X58 Classified
Intel i7 975
Corsair Dominator 1600mhz 3x2gb
Nvidia GTX 295
Auzentech X-Fi Forte
#8
KMoore4318
Pistoj Doulos Unum Pluribus
  • Total Posts : 17850
  • Reward points : 0
  • Joined: 2008/12/04 04:20:57
  • Location: Atlanta, Ga
  • Status: offline
  • Ribbons : 238
Re:I asked EVGA and Highpoint collaborate to fix this bizarre incompatibility. 2011/04/29 12:59:22 (permalink)
Asus used a newer release of the bios, our version had more limatations, the hole point of the 80 series of bios was to try to corect this isue, there was a long thread on it, you might be able to find it with a search.

  
Just clink Link to Register >  My Affiliate Code:VTJPDC4OMB 
1)965 V-8  E761(77)
2)980X AX1200 E760(77) 980 ti
3) E770 (77)2(590) SLI AX1200 
4) X299 FTW K, 9980XE 2 2080 supers on AX1200
#9
jmkuchta
New Member
  • Total Posts : 11
  • Reward points : 0
  • Joined: 2008/06/20 11:17:46
  • Status: offline
  • Ribbons : 0
Re:I asked EVGA and Highpoint collaborate to fix this bizarre incompatibility. 2018/12/05 23:47:22 (permalink)
Thanks for the info, y'all. Time ultimately solved this problem; today with NVMe drives, there's no longer any need to have RAID cards in any of my rigs anymore. And truthfully, UEFI BIOSes are far better than what we had before.
#10
Jump to:
  • Back to Mobile