Quote Originally Posted by Main View Post
Holy grief just ordered mine from performancepcs yesterday, and I have EX58-EXTREME
I live in Norway so it will cost a fortune to return it !
I agree. But let me take this opportunity to elaborate a little further here:

We have been working double time on this issue in the past 24 hours. For the record, when we designed our block, we used Intel official specifications (362033_362033_Nehalem_EP_EMTS_Ver1.52). If there are any revisions to this document affecting the keep-out zone, I am not aware of them.

The keep-out zone around the socket specifies that NO COMPONENTS SHOULD EXCEED 7MM IN HEIGHT ABOVE BOARD.

Knowing that the CPU IHS is 7.8mm above board, this obviously makes sense.

Gigabyte capacitors have been measured at 8.6mm above board. So they clearly are in violation of Intel specs. Now, are they the only ones? No Sirree.. Measured the caps on an Asus board, and the height is comprised between 8.7 and 8.8mm above board. The big difference between the Asus and the gigabyte boards is that Asus in their infinite wisdom, chose to place these caps far away enough from the socket so that they don't interfere with extra large heatsinks.

Note: any Intel rep reading this, welcome to comment or PM me.

/Rant on: So now, I am seriously frustrated because we have to revise several hundred base plates. Where are we going if M/B manufacturers don't respect Intel specs???? What's even more frustrating, is that I understand that [another] waterblock manufacturer is suffering the same pligh, but didn't do anything about it, forcing consummers to install their product in a less than optimal configuration.. /Rant off

Anyways, as far as you are concerned, this is what we'll do: send you a replacement plate without you having to return the "bad" one. Please contact michelle@swiftech.com to give her your address.

Quote Originally Posted by rge View Post
Mine is coming in tomorrow or Monday.

I have GB x58 EXtreme as well. Same issue occurs with vertical mount of HK waterblock. Vapor found if you nudge caps out of way, which I did for a while, it works...but I got tired of having to do that playing with different tims....so 3 minutes with a file, just rounded/flattened that one square edge, no more problems. Also with my finger had gently but permanently nudged the caps over by probably 0.1 or 0.2 mm as well.

If its the same issue as heatkiller, Ill be filing mine down. I appreciate Swiftechs great customer support...but my laziness for returning items knows no bounds.

If all your doing is rounding one edge, your obviously not going to affect the performance of the block, unless somehow you managed to make it leak/damage it, though dont see how if similar to HK issue.

It is this interfering edge on HK, that I rounded/flattened. NOTE: I am not suggesting anyone else do this, especially since Gabe is going to take care of it.


BTW...any chance of seeing a pic or description of swiftech base change...ie, is the edge simply being rounded/flattened?
I really do not recommend you do that. contrary to my earlier announcement, where we were planing on cutting a simple chamfer, it appears (subject to final validation) that the "proper" and final revision necessitates milling a 1.3mm deep x 2.5mm wide step all around the plate (avoiding the screws so as not to damage the threads). 7.8 + 1.3 = 9.1mm Plenty of clearance, even with lapped processors! But, it's not something you can do at home. If you butcher your plate, I guarantee you that you will lose performance, and likely not get sufficient clearance with the caps.

just email Michelle@swiftech.com, and we'll send you a replacement plate next week. However, since you are in the US, we will ask for your original plate back ;-)

Warning: the above discussion and information is subject to change, until final validation.