Jump to content

Official LanParty ICFX3200-T2R/G Overclocking Help/Discussion


Recommended Posts

  • Replies 96
  • Created
  • Last Reply

Top Posters In This Topic

The one thing anandtech was saying in the preview was they needed better cooling on the northbridge at 400+ fsb, and it was extremely difficult (as in too difficult for them to attempt it and risk loosing the board) to replace the stock heatsink, so they resorted to mounting a fan on top of it.


"The RD600 is cooled by a large passive heatsink that was well suited for the job until we started overclocking the FSB past 450. Due to the Northbridge voltage required to reach the maximum 511 FSB, the heatsink will require airflow over it. We would have preferred to replace the heatsink during this type of overclocking but DFI utilized a four hook attachment system that left us very little choice other than to direct airflow over the stock unit with an extra fan."

Share this post

Link to post
Share on other sites

That has to be crap. It's easy to remove the stock heatsink.. it looks no different than the one on the ASUS P5W DH, except for the extra clips. Push down evenly on top left and lower right clips and move them out, push down evenly on top right and lower left, and take it off.

Share this post

Link to post
Share on other sites

Yeah, upon closer inspection of the pictures, and the writing, I was thinking along the same lines-getting the heatsink off won't be a big deal. The real problem is there are no after market chipset coolers I am aware of that will clip into that spot, at least not yet anyway. I guess we'll see when street users have the board...

Share this post

Link to post
Share on other sites

That has to be crap. It's easy to remove the stock heatsink.. it looks no different than the one on the ASUS P5W DH, except for the extra clips. Push down evenly on top left and lower right clips and move them out, push down evenly on top right and lower left, and take it off.


the mounts are different than standard Intel 975 hooks


there are no heatsinks at this time that will fit this motherboard's NB mount unless you use adhesive

Share this post

Link to post
Share on other sites

Well, I think I can post the summary that I made before. Travis or others if you think there is some information that is not related in this thread feel free to chop chop :)


Most of the information below is coming directly from Tony (OCZ) and Bingo13 (who also have the board in his hands). So, please do NOT ask me anything cuz i am also like you at this stage.




Hey all,

After I spent 2 hours of reading and taking notes I wanted to share it with you also.

NONE OF THE BELOW INFORMATION BELONGS TO ME, but rather a summary of KEY points IMO. use it wisely!

Big thanks to Tony and Bingo!!




The manual is available for download at DFI.

3 PCI Express x16 slots for ATI CrossFire and ATI Physics

2+1 Configuration (2 Graphics + 1 Physics)

- 2 ATI CrossFire graphics cards (each operates at x8 bandwidth).

- 1 ATI Physics graphics card operates at x2 bandwidth.

3 PCI slots


2) Two things about this board make me think voltages will be more stable than what we've been used to on the Intel side...

a. It's got 6-phase digital pwm.

b. It's not an Intel chipset.


3) Its 16X with one card or 8x8 with 2, issue is just like Nf4 its fast with dual cards, I actually feel 16X 16X is not needed.


4) RD600 is actually quite old, I have a board here from over a year ago using a very early version of the chipset.

As far as i know there are 14 revisions of the chipset now, 13 and 14 being the best.

So....this chipset was actually designed ways back with RD480 which was 8 x 8 also.

EDIT: AMD just corrected me, 4 revisions of RD600, 13 and 14 being the 3rd and 4th...so 4 respins overall.


5) Bios shots: http://www.thetechrepository.com/showpost....139&postcount=6


6) of interest

DDR2 Ram 1.8V –range not shown

CPU VTT 1.2V – range 1.20-1.60V

NB Core 1.2V – range not shown

NB PLL 1.8V – range 1.80-2.50V

NB PLL 2.2V – range 1.20-1.45V

NB FSB Strapping Auto, 100,133,166,200,266,300, and 333

Address/Command Timing Auto 1T, 2T, and 3T




1) Just so you know.

FSB CPU 511 max thru bios

MEM FSB 658 Thru bios, so 1316 max DDR2


2) Droop

Idle 1.596

load 1.590

under nature 3d01, Orthos is crashing the board here. (feels like a code issue) All 3d at this time has been fine.


3) guys apps can have code issues with chipsets, we saw the same with early builds of memtest86 and prime had issues with Athlons also.....so it just may need a tweak or two to run 100% on this ATI chipset. Then again it may be just plain unstable...I will try to run some tests to see soon. (ORTHOS versus application stability talking) >>>> RAJU has a thread in off-topic section about ORTHOS, check it out!


4) There are not any memory dividers on the RD600 per say. The memory clock is true async if you want it be. I can set DDR2-1103 with a 283FSB or DDR2-981 if I want as an example. The memory clock is totally independent of the CPU FSB in async mode and it works very well.


5) Guys some more info for you;

RD600 is not as fast as i975 for superPI, if this matters to you and you spend all your life watching the PC crunch 20 loops then this is not the board for you.

RD600 is FAST in 3d though, and I hear with a hacked SLI driver its faster than i680 in SLI with 7900's.

So PI freaks may be disappointed, 3D benchers may be thrilled....you just need to decide what’s for you.


6) Ram performance @ 550 and 571: http://www.xtremesystems.org/forums/showpo...7&postcount=456


7) Sata performance: http://www.xtremesystems.org/forums/showpo...2&postcount=474


8) small summary:

a) 511 FSB at least 4 hours orthos stable (blend)

B) SLI hacked drivers work

c) Faster in 3D than 680i

d) memory clock is true async

e) up to 1.94 Vcore available

f) vdroop of .006 @ load seen thus far


9) Tony's guide: http://www.thetechrepository.com/showthread.php?t=76


AVAILABILITY (from Tony's 3 days ago post)

1) Who did you speak with?

I just got off a call with DFI USA in an attempt to settle this argument of will there or won't there be boards. This is what I know.

They have a sample run made, these boards should hit the USA before Xmas, this is like the 975 infinity where 30 boards or so shipped.

Now...there are some tweaks to be made, this was the whole point of 4 people getting the board to test. These should be just bios issues though so no hold up is expected on this first small shipment.

Final run will be made soon and these boards will be available in the new year, this was always the case for absolute mass production boards and not the pilot run.

I hope this settles the will there or the won't there be...all being well there will be a small shipment as i have been saying all along, if something serious turns up though there could be a small delay which would push things back to the new year.


SOURCE (especially big thanks to Tony and Bingo):
















Share this post

Link to post
Share on other sites

so, Travis - where are you in your testing?


up n running at stock - any overclocking attempts - already fried the first one, hehe!


I am little bit disappointed from early reports that fsb is limited to 511mhz. It's great for 6600/6700 users, however for minority 8x511 simply won't cut :)

Share this post

Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in

Sign In Now

  • Create New...