I have read the Nforce threads about running 4 gig sticks in board recommended for only 2. I also just read a thread in this section talking about the fact that it may be possible or may be possible soon.
I am wondering if any one has tried it. I just built a computer for work. Solidworks is the main resource intensive program I will be using. I think 16 gigs will be more than enough for now but now that I have started thinking about it and I could pick up 2x8 for about the same price as 4x4 I'm wondering if it might be possible.
Has anyone tried it?

wait until this board gets the BIOS microcode update for Ivy Bridge, then I'd say it should run well with 8GB sticks (aka max 32GB RAM). current BIOS is old, it may (or may not) run with 2x 8GB sticks, up to you to test it.

Any idea when the Ivy Bridge update will be?
I stole one of the 4 GB sticks from my home computer and that will probably get me by for a little while. I suppose I could just order the 8GB and hope they will work after the update if not before.

I would tend to go with Gskill since that's what I currently have and it seems like good memory.
If I did try this though should I go with the 1600 with low timings http://www.newegg.com/Pro...x?Item=N82E16820231568
Or 1866 with higher timings?
http://www.newegg.com/Pro...x?Item=N82E16820231528
Price is an important factor in the decision.
Otherwise I'd just jump up to these.
http://www.newegg.com/Pro...x?Item=N82E16820231571

I've tried with 8 GB for 1week and no problem at all.
Got 16GB (Look at SIG)

So I'm thinking about going with the 1600 sticks unless someone jumps in and tells me I should do otherwise.
 

Use the 1.5v memory sticks if you can. It is what is recommended for the SB and IB CPU Chips.  I know others use the 1.65v... but why? More heat from the memory sticks, more heat/on the cpu chip itself to to the higher VCCIO required) to get the drive required for more memory.  I alway try for Corsair memory first then the G.Skill.  Above is just my experience ymmv

Thanks. Both I am looking at are 1.5 v. I am just wondering if there would be any significant advantage to the 1866 over the 1600.
Most of what I have seen shows pretty minimal gains between the two and the lower timings of the 1600 makes it seem like it would be a decent option.
Just want to make sure I'm not missing anything or would be better off with the 1866.
What are the chances of being able to OC the 1600 to 1866?

Hmm never knew you could use 8GeeBee sticks

Got the memory in today. These are for my work computer which is a z68 SLI micro running an I3 but I decided to throw them into my home desktop to try them out tonight.
http://www.newegg.com/Pro...x?Item=N82E16820231568
Z68 SLI and 2600K.
stock timings are 9-9-9-24-2. I started it up at the rated 1600 and it seemed fine so I restarted and turned it up to 1866 which is what I was running my Gskill 2x4 2133 at. It started up and ran fine. Still running the same timing and 1.5 v.
Test has been running for 15 min so far just fine.
 

 

It shrank the image too much to see lol. Let me blow it up a little bit
 

Ok weird.
I loaded a previous OC profile in the bios and in that profile I had the mem voltage set at 1580 mv but when I loaded that profile and restarted the bios it said mem was 1500.
Linx stopped with an error at 38 min. When I restarted the bios said memory was 1580 mv so I'm not sure what it was running at before. Anyway I'll restart linx and see what happens.
I don't really want to go over 1.58 v so if it continues unstable would it be better to drop the frequency or should I drop the timings and if so which ones?
The other issue is I'm running Linx under Wine in Linux so who knows if that could cause it's own glitches.
My work machine will be running Windows 7 though so I can play with it more then.

Yep error after 20 min this time.
What should I do next?

VCCIO. crank that bad boy up. 

VCCIO is currently in auto @ 1.05 v. What's safe to take this to?
 

I believe its up to 1.20 is perfectly safe.  you can google around for the z68 safe limits. 

try 1.2V (it's safe) on VCCIO and re-test. you shouldn't need to up voltage on VDIMM if running them at rated speeds. you may need a lower VCCIO than 1.2v, but try that first and see if it helps any, you can optimize later.
 
maybe try running memtest86+ 4.20, it's been designed for picking up slight memory errors. if they pass one pass then it's probably good, but eventually always double check the final settings with an overnight run (several passes one after another).

1.1 v stopped on error. 1.2 stopped on error.
Is there any concern running 1600 ram at 1866?