890FXA-GD65 and possible USB Controller related issue?

Question

Hi all - Been having a bad time of it with a new system build for a little under a month now.  Have tried to capture my "notes" below, any idea of a root cause?  Is anyone aware of known issues with the board; the board and Windows 7; the board and some other component(s)?
My issue:  During Windows 7 x64 installation or startup, the system will freeze at the "Starting Windows" screen.
My system specifications:
MOTHERBOARD: MSI 890FXA-GD65  (BIOS ver 18)
CPU: AMD Phenom II X2 555 @ stock speed  (HDZ555WFGMBOX)
MEMORY: DDR3 1600 2X4GB G.Skill Ripjaws-X [DIMM1 & DIMM2] (F3-12800CL9D-8GBXL)
VIDEO CARD: XFX Radeon HD 5670 512MB (HD-567X-YNF3)
HARD DRIVE(s): 2 x Western Digital 500GB 7200 drive (WD5000AAVS)
POWER SUPPLY: Antec NEO ECO 620C 620W
Windows 7 x64 (Ultimate edition, retail)
This problem has existed since I first built the system and is consistently reproducable.
I am able to avoid the issue by disabling the USB Controller in BIOS prior to initiating the Windows installer.  This will allow me to proceed through Windows installation, but as soon as the USB Controller is enabled once again, the issue recurs.  Having to disable USB support is undesirable.
I'm now on my second 890FXA-GD65 motherboard, having RMA'ed the first board after a fairly brief exchange with MSI tech support(2 replies, with the eventual suggestion of RMA for a faulty USB controller).  So I went through the RMA process with the online retailer, receiving a new board today.
The problem is that the RMA replacement behaves in exactly the same manner.  Normally this would suggest to me that the problem may lie with one or more of the other components, but I've not been able to identify issues with any of them.  One obvious suggestion would be to try swapping out components, but my resources are constrained on that front - video card happens to be the only component for which I have a alternate that I can test with, actually.
As far as troubleshooting investigation, here's what I've done/observed:
1. Windows installation/startup will succeed if the USB Controller is disabled in BIOS.
2. The same behavior can be observed with a Windows Vista installation disk as well.
3. When loading Windows in safe mode, the last driver file loaded was disk.sys
4. Memtest86+ (v4.2), repeated full passes reveal no issues with the system memory.
5. While seeming that it would be an unlikely factor, I have swapped out video cards(to an eVGA 8800 GTS 320MB), but observed the same issue still exists.
6. Running in a single DIMM configuration affected no change.
7. Have disconnected one or the other of the hard drives with no change.
8. The BIOS version installed is equivalent to the latest(and only) version posted on the MSI support site.
So...any ideas on what's left to try?  I've done enough reading online to notice that I'm not the only person to have observed this behavior, but have yet to encounter any explanation of it nor a resolution.
Any suggestions?

Answer 1

Trying to catch the points that weren't explicitly spoken to by me in the last attempt:
@Stu - Regarding testing with zero USB devices connected - this was done as part of troubleshooting and it has no observable effect, so I think that I can rule out a bad USB device being the culprit and I apologize for not making that clear previously.
@Frederik - I've done a little more legwork on the memory item.  Troubleshooting, for the most part, has been executed with this memory set to run as DDR1333 @ 1.5V as that is what the board auto-defaults to(the kit is rated as DDR1600 @ 1.5V)  It was at the defaulted setting that I ran memtest86+ and no errors were reported.
I've also just now made an attempt to run this memory as DDR1066 at 1.65V and still there is no change in results.  In my research on this specific kit, I have not run into accounts like mine - in the accounts that I have read regarding a poor experience with the kit, memory has either been been obviously bad, for which failures are reported in diagnostics like memtest. ...or, I suppose that there's also the random guy complaining about his overclock not working or the memory speed defaulting to the wrong value in BIOS. 
Frankly, I don't want to battle any particular component and I owe no allegiance to this make of memory.  I do want to try to be objective, given that I'm stuck with these modules until I make the decision to RMA.  As such, I had read what info is available out there - prior to purchase, I hadn't noted so many reports of issues with this kit across a considerable sample size of feedback so as to cause me initial concern, or else I would have gone with a different kit/different brand.  Probably will do that anyways, next time, but in the meantime, I'm afraid of this being a distraction.  If both myself and this other person(drgan) had both been using the same memory kit, and thus it was more likely to be the common thread, I'd certainly be more inclined to this line of reasoning.  But since he's using Kingston and in the few other accounts of a system freezing on USB controller enabled status I've noted varied memory kit brands, it doesn't seem to me to be an obvious culprit.  But if not, then what's left to consider?  The other interesting factor to me is that the issue appears to be Windows-centric....but...