odd problem with new build

[Matthew]

I just put together the following system:

Win 7 Home-64
MSI X58 Pro-E LGA 1366 Intel X58 ATX Intel Motherboard
Kingston Triple channel (2GBx3) PC3-10600 (MB approved)
Intel i7-930 2.8GHz processor
Western Digital Caviar Blue 320GB system drive
Western Digital Caviar Green 1.5TB data drive
Sapphire Radeon HD 5450 video card
Cooler Master 650w power supply
Samsung DVD/RW drive

I went ahead and turned RAID on (for future use) before I installed Win7,
but just use the one system drive and the one data drive at the moment. I
also installed the Intel drive management software along with all the other
latest drivers and BIOS for the motherboard.

When I boot the computer it hangs at the post screen saying "detecting AHCI"
for about 3 minutes then boots... and when I hook up any sort of USB hard
drive to the system it takes roughly 5 minutes for the system to recognize
the drive, even if it has detected it in the past. I built the system for
Photoshop work and if I try to load more than a couple of pictures at a time
the system becomes unresponsive and Photoshop crashes. If I just open up
Photoshop and then try to run a second application the system hangs again.
And if I do manage to get several pictures open at once, if I run any sort
of batch command on them the system hangs again.

Is it possible that initializing RAID or possibly this Intel drive
management software is effecting performance of my drives? I'm assuming that
the initial start up hang at "detecting AHCI" is also somehow related to the
RAID configuration, but I see no option for disabling AHCI in the bios (if
that's even possible). Or could there be something else I am missing?

 

[ToolPackinMama]

Is it possible that initializing RAID or possibly this Intel drive
management software is effecting performance of my drives? I'm assuming that
the initial start up hang at "detecting AHCI" is also somehow related to the
RAID configuration, but I see no option for disabling AHCI in the bios (if
that's even possible). Or could there be something else I am missing?

Have you tried booting with BIOS setup defaults?

 

[Matthew]

Have you tried booting with BIOS setup defaults?

well... what would that do if I've installed Win7 with RAID? Or will it
matter since I'm not running RAID at the moment? That is pretty much the
only change I made in the BIOS.

 

[Pen]

well... what would that do if I've installed Win7 with RAID? Or will it
matter since I'm not running RAID at the moment? That is pretty much the
only change I made in the BIOS.

Turning on the RAID turns on the AHCI which is required by
SATA drives. All the rest of your problems scream Memory
troubles with the system swapping out to the hard drive.

 

[Paul]

Turning on the RAID turns on the AHCI which is required by
SATA drives. All the rest of your problems scream Memory
troubles with the system swapping out to the hard drive.

In addition, Matthew might try another SATA cable, as if the
SATA cable was bad, and causing a lot of CRC errors, that
might result in the drive being accessed over and over
again. Perhaps the Southbridge SATA interface is in a
reset loop, trying over and over again to establish
communications ?

The hard drive manufacturers sometimes have downloadable
diagnostics, and it would be interesting to see whether
the diagnostic can make sense of it or not. The Western Digital
product page for each drive, should point to the diagnostic
to use. For example, I can see "Data Lifeguard Diagnostic" here.

http://support.wdc.com/product/download.asp?groupid=606&lang=en

A memory test, should always be done on a new system, before
trusting it to do anything properly. Scroll half way down the
page here, for some downloads. Some Linux LiveCDs, include
a boot command prompt option, of doing memtest as well. So
it's even being distributed on some Linux CDs.

http://www.memtest.org

LGA1366 systems sometimes have problems with socket contact
to the processor. It can affect how many sticks of RAM are
detected (some people "find a channel missing"). Check the
BIOS to see if all the memory is indicated as being present.
The memory must be detected, before a memtest program can
test it.

Paul

 

[Matthew]

In addition, Matthew might try another SATA cable, as if the
SATA cable was bad, and causing a lot of CRC errors, that
might result in the drive being accessed over and over
again. Perhaps the Southbridge SATA interface is in a
reset loop, trying over and over again to establish
communications ?

The hard drive manufacturers sometimes have downloadable
diagnostics, and it would be interesting to see whether
the diagnostic can make sense of it or not. The Western Digital
product page for each drive, should point to the diagnostic
to use. For example, I can see "Data Lifeguard Diagnostic" here.

http://support.wdc.com/product/download.asp?groupid=606&lang=en

A memory test, should always be done on a new system, before
trusting it to do anything properly. Scroll half way down the
page here, for some downloads. Some Linux LiveCDs, include
a boot command prompt option, of doing memtest as well. So
it's even being distributed on some Linux CDs.

http://www.memtest.org

LGA1366 systems sometimes have problems with socket contact
to the processor. It can affect how many sticks of RAM are
detected (some people "find a channel missing"). Check the
BIOS to see if all the memory is indicated as being present.
The memory must be detected, before a memtest program can
test it.

Paul

It's odd that you mention memory because originally I had two kits of
Kingston Triple channel (2GBx3) PC3-10600 (MB approved) memory. The BIOS
would detect all 6 sticks but only report 8GB available memory, so I tried
each 2GBx3 kit individually and one kit showed up in the BIOS as 3
sticks/6GB available and the other one showed up as 3 sticks/2GM available
so I RMA'd it back to the store. I assumed that since the BIOS detected 3
sticks at 6GB that this memory was good to go, but I'll test it out tonight
with memtest. The odd thing is that I pulled this memory strait from MSI's
list of compatible memory for my board.

 

[Paul]

It's odd that you mention memory because originally I had two kits of
Kingston Triple channel (2GBx3) PC3-10600 (MB approved) memory. The BIOS
would detect all 6 sticks but only report 8GB available memory, so I tried
each 2GBx3 kit individually and one kit showed up in the BIOS as 3
sticks/6GB available and the other one showed up as 3 sticks/2GM available
so I RMA'd it back to the store. I assumed that since the BIOS detected 3
sticks at 6GB that this memory was good to go, but I'll test it out tonight
with memtest. The odd thing is that I pulled this memory strait from MSI's
list of compatible memory for my board.

As long as all your memory is being detected, there I wouldn't worry
about it. If you still had six sticks, you could experiment with
reseating the processor, to see if all of it would show up in the BIOS.
That is, on the theory, that some contacts on the LGA1366 aren't making
connections. The memory terminates on the processor, so either
the DIMM slot could be bad, or the processor contacts could be
open circuit. (Processors are thoroughly tested, so I wouldn't
suspect a bad processor, at least initially. A contact problem is
more likely.)

You could also take a single stick of RAM, and test it in each DIMM slot,
and see if the problem is only with certain slots.

Foxconn had issues with LGA1156 sockets. I don't know who is making
sockets for LGA1366, and I haven't seen any public discussions of
LGA1366 issues, like there was for LGA1156.

http://www.anandtech.com/show/2859

Paul

 

[Matthew]

As long as all your memory is being detected, there I wouldn't worry
about it. If you still had six sticks, you could experiment with
reseating the processor, to see if all of it would show up in the BIOS.
That is, on the theory, that some contacts on the LGA1366 aren't making
connections. The memory terminates on the processor, so either
the DIMM slot could be bad, or the processor contacts could be
open circuit. (Processors are thoroughly tested, so I wouldn't
suspect a bad processor, at least initially. A contact problem is
more likely.)

You could also take a single stick of RAM, and test it in each DIMM slot,
and see if the problem is only with certain slots.

Foxconn had issues with LGA1156 sockets. I don't know who is making
sockets for LGA1366, and I haven't seen any public discussions of
LGA1366 issues, like there was for LGA1156.

http://www.anandtech.com/show/2859

Paul

Well... if anyone cares... with RAID on USB harddrives and thumbdrives
basically do not work, turn RAID off and everything works just fine, turn it
back on again and nothing but problems... guess that means no RAID for this
system.