Does anybody that has this board use phase, i've been looking for the 500fsb screens and havent seen anything yet. Im more interested in 9x multi results.
Does anybody that has this board use phase, i've been looking for the 500fsb screens and havent seen anything yet. Im more interested in 9x multi results.
Thanks for the heads up on the connectors, I skipped the 1x4 and second hd led , the power led doesn't fit so woopdido.
When OC'ing is the jumper still a big part of screwing up or have me moved past that yet?
nice one doug. 59c Load is far from being harmfull. If you load both cores 100% 24/7 (folding at home) it would be a sweet spot but if you only load sometime during the day, you can load up to 70c.... Photoshop wont put as much load as orthos does, maybe your simulation stuff but its pretty rare that a program use 100% both core for more than 2 seconds eheh... You could go 3.4ghz, try and see if it needs more voltage. Just put a little bit more if it needs and it will be your max with stock cooler.
But Yes the motherboard seems to be really stable, so far im at the same point where supertimor started this thread. 3800mhz Orthos stable, 1.3vfsb, 1.6vmch (its low~!) 1.465vcore. Load @ 69c but im not 100% sure coretemps is really right... Remind you guys that this is a L631B chip so it runs hotter.
http://www.oliviercouture.com/PC/Ben...table_3800.JPG
I am aiming for 3850mhz stable and 4000mhz benchable but 4ghz seems to be really hard for air cooling with a B chip..... @ 1.5vcore i load over 75c ..... then again if someone could prove that coretemp is not really accurate at a certain point it would help me pushing it harder :P
unless you get really lucky, you probably won't see much over 470. this board works best at sub 400 FSB.Quote:
Originally Posted by MrDeeds
I thought 400-450 range was common for this board if you used the 1333 strap...but you seem to indicate sub 400(is this on a quad core)Quote:
Originally Posted by Blauhung
hehe, it works best sub 400fsb, but it will max out in the 450 range on the 1333 strap with the very relaxed timings that follow with the strap change. I simply meant that the board is tuned to have the best performance at sub 400 FSB.Quote:
Originally Posted by nealh
Just trying to steer people that are coming here looking for max FSB off to the 965/6xx(nvid/ati) boards where they will be much happier.
edit: to show what i mean, currently my best performance is at FSB300 on the 800 strap. This is way better then higher FSB and higher clock and memory speeds using lower multi's and dividers simply because of the performance hit you take due to relaxing your timings through the strap change.
CPU 300x12
ref freq 200
mem freq 667
DDR2-1000 4-4-4-15
i'll put up some shots of pi and 3do6 when i get a chance to play this weekend
well im at 422 on the 1066strap (266 ref frequency) I will continue to push it but i dont know what is the limmit of the 1066 strap. I dont know what issue you have when youre strap is too tight, is it like a normal unstable system: random shutdown/orthos fail after x seconds ? Right now my system is 100% stable at 422x8 3800mhz 1.3vfsb, 1.6vmch 1.465vcore, 266 ref freq, 533 mem freq (1:1) 2.16vdimm. I am thinking that 3850mhz on the same vcore could be easy but it looks like its far from being stable, maybe its because of the strap ? i may want to go on 1333 but at least if i can get more bench performance (i.e higher clocks but higher chipset latency but still better performance).
Also im sticking to 1:1 right now (844 mhz 4-4-3-4 trc 7 but i know teamgroup d9gmh can do higher but im not sure if other divider is worth the time to max out, because alot of people says 1:1 is better on 975x ....
It would be really nice to have a comparison chart between 800/1066/1333 strap for Spi / 3dmark / Orthos stability / Real World Gaming Demos...
I'm working on that chart. I finaly just got my 8800GTX after stupid UPS briefly lost the package so now i can do 3d benching as well (nforce 6800 didn't cut it)Quote:
Originally Posted by shogo_ca
but i can tell you that moving along the different dividers does alter the extended latency settings just like the strap change did. attached is a bit more extended table. You can see that bumping up the strap did raise some of the timings back to where they were on the 1333 strap, but the overall bandwidth and latency went down because of the increase in ram speed. I'll get the other dividers for the higher straps this weekend.
ah yeah i remember seeing that thing, it is really nice. I takes some time and effort to take a deep dive into strap/divider/fsb performance. Cant wait to see 3dmark 06 / SuperPI
Gents, Whats the take-away from this chart? For CPU intensive calc's memory bandwidth is large factor in performance. I assume this is mainly true for cases where lots of external mem access is required. How does this translate to real-world applications?
See for me I am trying to decide betweeen this mobo and waiting for RD600
I have an e6600 I bought that will run at 3600 maybe not orthos stable ..not sure
SO I want 400fsb that is stable and easy to achieve
My ram is 10th anniv crucial ...supposedly will do 1000 i many cases at 5-5-5-15..all I want is 800@4-4-4-12
It all depends on what you want to do with it as far as real world apps. I just wanted to start showing what i could of how the board alters latency with straps and dividers. In the bulk of apps run on a C2duo you don't really need as high of a bandwidth, its much more about latency (mostly to help out with pre-fetching tasks).Quote:
Originally Posted by doug goldberg
keep in mind that for me my goal is to keep latency low at high cpu clocks and high memory clocks. I can do all this with an unlocked chip and low FSB. anyone with a E6400 or lower (maybe a 6600) gets to the point where to get high cpu clocks they have to use the FSB to push their chip up and that means you have to relax the strap and give up performance. I'm going to try and run through lower multipliers on my chip and replicate what 67/6600 users could expect if they run up the FSB on each strap, but it really takes time and girlfriend is more fun and work buys more goodies for me to play with.
From the sounds of things
800 strap = good to ~330FSB
1066 strap = good to ~380FSB
1333 strap = good to ~450FSB
As i was saying, im running 1066 strap at 422fsb, 100% stable. Maybe it could depend on the board revision ? i have 503 ..Quote:
Originally Posted by Blauhung
Anyway further testings are needed to have a conclusion on this. I will try to push on the 1066 strap as much as i can and then move on to 1333 and test in 3dmark to see performance difference.
I just received my board last night, got a rev 505 board from Tigerdirect (thru Amazon for cheap shipping :) ) I couldn't wait for DFI's RD600 board any longer, and this board looked like just what I need. So far I am impressed with the board, the DFI will have to do everything but make me breakfast to pull me away from this board ;p I haven't had time to play around with overclock but I was able to hit 3.3 very easily after about 20 min of tinkering and a couple failed boots. I didn't have a clue how to reset the board after a failed OC, guess just pulling power works?
Quick question, any problems with having both IDU and IDCC installed? I installed em both at first trying to figure out which one is intended for this mobo but they both work lol.
to reset the board you will have to: power off, unplug cable, put the jumper in from pin 1-2 to 2-3, boot in that position, correct the settings, put back in 1-2 position, boot, voila.
Or you can also have a switch like Supertim0r have !
Yes This board seems more stable. I have used p5w dh and p5b deluxe with 6600s and 6400s and they were not booting as easy as badaxe2 with quick overclock/no testing
Just installed the quad core...damn its a HOT chip :eek:
OK here goes my stoopid question
I don't use Creative products and don't care to spend on a Dolby digital card either
So how good is the on board sound ?
From a scale of 1 to 10
10 being the top after market sound cards and 7 being the nF2's Sound Storm,the best on board sound ever made.
couldn't say, i'm using a X-Fi or X-MERIDIAN
i was only witting down the averages, give or take 20 or so MHz depending on luck of the drawQuote:
Originally Posted by shogo_ca
I got my latest BX2 up and running tonight. The first one I had was a dud, it froze during the first attempt to enter the BIOS config and I never got it to fire up after that. This one is FAR better and seems to be working great.
It's a rev 505 with the 2333 BIOS out of the box.
I am running an E6600 at 9x400, 1:1 on 1066 strap, 4-4-4-12 @ 2.2v (PDP 6400LL). vfsb/vmch are both stock, but the MCH heatsink is scorching hot (anyone else notice that?). The C2D needs 1.55v (1.6 selected in BIOS) at this speed, but it's on water, so with Orthos fully loading the cores, it's staying around 60C. I'd love to run 1.4-1.45v like some are able to do, but this chip needs the juice. Maybe I'll get a better one later...
So far I am loving this board. Highest stable on three other boards (965, 680i & 975) was 3.4, and I could barely get 3.5 stable with a lot of vcore thrown at it. The 965 would inconsistently boot 3.6, but always locked at the desktop. Keeping my fingers crossed to keep 3.6GHz stable on the BX2. :woot:
Just my $.02 contribution to the thread. Carry on. :D
Tom
Alright mines here, how do I pop this stupid NB sink off?
Your not just whistling Dixie, I called Intel for the hell of it and asked them what the idle temps should be and they say 60C:eek: So I'm @ 53C or so but with no OC. Its funny when I look at the intel desktop utilities and check cpu temp its always in the yellow caution area:DQuote:
Originally Posted by Supertim0r
you have to remove the black plastic clip first iircQuote:
Originally Posted by afireinside
lets us know how it OC :)
53c no OC with Tower120 ???Quote:
Originally Posted by palese
I can prime @ 3.5ghz 1.425v in bios
70-66-69-66c
2395 bios or 2333 bios? Thanks.Quote:
Originally Posted by shogo_ca