**Official DFI LanParty UT P35-T2R Review/Overclock/Guide Thread**

[Praz]

Most any board, server class or otherwise, has the probability of the above mentioned squeal. The reason we as enthusiasts see this so much is because of how hard we push the hardware and the options available that allow this push. The only real solution is higher grade and/or additional components. It's doubtful the market would bear the price point that would be required for the components that would have to be used. The second solution of additional components would be less of a cost issue but the needed real estate on the board doesn't exist.

[Brother Esau]

Good morning, John nice to see you Like I said its Intel's fault and we are not asking these motherboards to do anything that a server stress environment does not create for any given hardware! Excuse me I do not push my stuff to out of control voltages like allot do in theory and in practice I find the max and back off 200-300 mhz Dependant upon how much voltage was needed to find my ceiling.
So in actual fact whats the problem on my end? I am very level headed when it comes to O.C and not unrealistic in the sense that I don't run electric chair voltage through my equipment and expect it to last and not to fail!

[RLM]

the whistling is the pwm/coil packs
it has nothing to do with memory, it has everything to do with stressing the board.
if its more predominant when running mem benches then that indicates that the coil around the ram area would be at fault, possibly. but not definately.

That's my ignorance showing Letsat. Possibly adding 2 and 2 and getting 3.

Had the same squeal on my Quad GT. It came from a coil just under the memory slots. However the machine ran stable as a rock and gave me months of 3d rendering without a hitch.

On the commandos I seem to remember people were coating the coils in silicone sealant to prevent the noise.

Although I did not inquire as to the specifics I do know that Quads were tested on this board at DFI.

I read a thread the other day, where a guy had written to DFI about problems with his 8500 memory. Allegedly DFI replied to him saying that they hadn't tested 8500 sticks with the board.

It was this response in part that fueled my speculation.

If they have been tested with Quads, and I hope run of the mill quads, then that's reassuring.

The last couple of weeks, along with running my own copious tests, I've spent a lot of time trawling forums, including babelfished Japananese, German and Italian ones. It's been difficult to find any decent info for running a Q6600.

What's needed or wouldn't hurt is some sort of guildelines. For instance as I say with my quad findings, too little voltage in the VTT seems to be just as detrimental as maybe too much. I want details Or is that maybe too much too ask.

EDT and EA set to Normal can result in as much as a 8-10ns loss in latency time depending on other settings.

Did a lot of tests with EA and EDT set to both Fast and Normal. Yes Normal is quite a bit slower. However on both my Geils and Gskill I could not run them at 1000+ cas 4 with fast settings, stably. That was irrespective of performance levels, trfc, clk, voltage or loosening other timings.

Not knocking DFI for this, I just put it patiently down to early bios.

Not trying to sew the bad seed here. I've had some decent results our of the DFI, and optimistically I'm hoping that my replacement board will be all good. Just hoping I don't have to wait too long

RLM

[Raja@ASUS]

I am very level headed when it comes to O.C and not unrealistic in the sense that I don't run electric chair voltage through my equipment and expect it to last and not to fail!

A wise choice for a 24/7 system, good voltages keep things alive. One of the worst things a person can do is run a system continuously on, or near benching voltages 24/7 and expect things to last, ram is usually the quickest thing to fail.

regards
Raja

[retrospooty]

Good morning, John nice to see you Like I said its Intel's fault and we are not asking these motherboards to do anything that a server stress environment does not create for any given hardware! Excuse me I do not push my stuff to out of control voltages like allot do in theory and in practice I find the max and back off 200-300 mhz Dependant upon how much voltage was needed to find my ceiling.
So in actual fact whats the problem on my end? I am very level headed when it comes to O.C and not unrealistic in the sense that I don't run electric chair voltage through my equipment and expect it to last and not to fail!

I am not so sure its an Intel deal... I have had the annoying high pitch squeal when heavily overclocking on several motherboards, including my AMD based DFI lp ut nf4 sli-d.

Now for the good news, I have a DFI LP UT P35-T2R (ugh, too many letters) and it does not appear to squeal under heavy OC load.

'least I'm enjoyin' the ride !