Printable View
Correct. The reference frequency tells the MCH what speed it should assume it's running at and to set its internal timings appropriately. The lower it's set, the lower the MCH latencies. Just like memory, there's a tradeoff between latency, speed, and voltage.Quote:
Originally Posted by xMrBunglex
ok, that makes sense, thank you.Quote:
Originally Posted by gtj
And so along that same line of thinking can I assume that 1:2 in 200 will hit a wall before 266, and 266 before 333, etc, as FSB increases? Is this overcome through increased vMCH?
and this all more or less goes back to my issue of not being able to effectively set my multi below 10, and therefore i am trying to maximize bandwidth in other ways.
A little bit can be overcome by voltage but not much. It's mostly FSB.Quote:
Originally Posted by Sean-E-Boy
You should report an issue to Intel on the multiplier issue. I'm going to report it as well.
http://supportmail.intel.com/scripts...d=36,2778,2807
issue reported.
and escalated
well, i dont think that my strategy is going to work:(
it seems that 200/800 tops out around 330FSB for me and 266/1066 is about the same. going to try a few other things, but it is probably back to 333/1333 for me until i can get this muliplier thing worked out.
anyone had any luck with GTL values and independent core stabilization?
Sean,
Have you tried enabling EIST under the power menu to see if that makes the multiplier setting work. It works for me if EIST is enabled.
yes, i tried that too...no workie.
i really don't know much about GTL values. i'd like to learn more though.
attached is the reply from Intel. I expected something like this, and hopefully they still keep track of the issue and pass it on to someone that can do something about it.Quote:
Originally Posted by Sean-E-Boy
However, I didn't expect that more than half of the reply would be about "please rate our reporting system..."
I have some really good documents related to the issue (but i hope that you have a PhD in electical engineering!). I will post them later today once I get home.Quote:
Originally Posted by xMrBunglex
as promised
<<attached>>
the intel one is more technical than the other
sean
Don't get caught up in the "overclocking" spiel. Reply back with a statement that "A BIOS feature is not working correctly". That should key them to pass the case on to the BIOS team. It takes a little patience (just like the board), but you CAN get to a live person who actually has a board in front of them and will try it out. They've gotten a little more "enthusiast friendly" but the autoresponder and tier 1 hasn't got a clue.Quote:
Originally Posted by Sean-E-Boy
Good finds. :)Quote:
Originally Posted by Sean-E-Boy
Thank you.Quote:
Originally Posted by gtj
And as per your recommendation, I have replied to Intel's reply.
Typical Intel reply, always oc is not supported, even if the matter is not directly related to oc. I guess it's always their first move to see if we just let it go...
Anyone knows about the C2 state option? I guess it's a new processor state but any details? C0 and C1 are known, this one must be new.
IS this board a good Overclocker With X3360 (Q9550) CPU ?
What would your goals be?Quote:
Originally Posted by X2beach
Has anyone tried a QX9770 on the DX38BT, since the bios for the X38 is currently the same as the X48 it should work right? And by working i mean will it detect the proper speeds automatically.
Just curious.
does this board allow you to tweak the GTL and anyone got q9550 to run 450 bus on this board? DFI I cannot get it to run more then 442bus
The motherboard will obviously run at a 1600 but whether it will be automatically detected I'm not sure.Quote:
Originally Posted by rchelicopter
The latest BIOS does allow you to tweak the GTL but I had an E6850 stable at 450x9 without GTL adjustments. Can't say on a Q9550 though.Quote:
Originally Posted by X2beach
I am thinking about getting the DX48BT2 since I have some OCZ DDr3-1600 just laying here I like to use 4gigs of it..
I wonder if this board is more stable the the DFI X48 ?
Quick update on my set-up.
I got some time to put everything on the bench and run memtest. It looks like I can get all 4GB running 1333 stable but I have to set the MCH @ 1.4.
Anything I should be concerned about running it that high ?
you'd actually be willing to buy a new motherboard just so you can get an extra 8 MHz FSB? you obviously have too much money - i'll be glad to take some off your hands :D i accept cash or PayPal.
the highest i ever tried to get my BT was 467 FSB (for an even 4.2 CPU) and it handled it easily with moderate volts and stock NB cooling with no fan. a former member of this thread hit 500 FSB (i think he's the only one who tried it). none of us have an X3360 though. i'm not sure how that would affect the results.
I got one on the way and the Asus Striker2 Extreme.
I will post results from both this weekend..
i've been running my vMCH at 1.425 and vFSB at 1.35 since i got the board. no NB fan, stock thermal pad on the heatsink... totally stock. my NB idles at 30° and i think i saw it as high as 43° once during an Orthos run in warm weather. you'll be fine at 1.4.Quote:
Originally Posted by Randy