G.skill RAM chip question

Does anyone know what type of chips this kit uses?

http://www.gskill.com/en/f1-3200phu2-2gns.html

It could be any of the Infenion IC's ranging from BE-5's to CE-6. My bet would be that they are CE-6.

They are all avg if you don't plan on OC'ing them much.

How do they like voltage? Would I be right in saying 2.9-3.0v is the max worth trying?

They are not a big fan of Vdimm or loose timing to tell you the truth.

You will push max out of them with 2.5-4-3-8 and 2.8V. Looser timings and higher Vdimm will hardly make a difference.

Thanks, I'm going to try that right now. 1T or 2T recommend?

Edit: 1T got errors. Testing 2T @ your config. Note: In my main system I got 2.5-3-3-6 @ 219 MHz stable. Now I'm running these in the Lanparty Expert board, and am testing 219 MHz as well @ the settings you gave me. I am going to try to maximize the frequency with those settings. Running 2T. Probably was run 2T in the ABIT board.

Edit 2: Exact same errors @ 2T. Wtf? Attempting 2.84 volts (supposedly was running @ 2.85 in the abit board).

Edit 3: FROZE @ higher voltage at same point used to get errors. DOH. Will get back in a while.

Yea don't expect much out of these. Mine did 220MHz MAX with secondary timings tweaked and 2.9Vdimm. They just wouldn't budge 1MHz more.

If I were you set to 1T try 210MHz with 2.5-3-3-8 at 2.8V and if it works then leave it at that and use dividers to OC your proc.

What I don't get is why doing 219 MHz is so difficult on a DFI Lanparty Expert, when on an ABIT IC7-G 219 MHz was rock solid at stock timing. Yes, I am referring to the same RAM sticks. Doing more testing with auto everything and 219 MHz + 2.8v.

It may have something to do with the added alpha timings.

You may want to go over to DFI-street.com and see if you could find the correct bios settings for those stick.

Well I've found some settings that allow them to pass memtest at 219 MHz but they ain't pretty! 3-4-4-8, a bunch of other crazy timings that I don't understand, and finally needed 2.96v to get no errors, or only one error, in test #8. One step down from that yields 2 errors. One step down again yields about 13 errors. Etc.

Thanks to the Xp120's 140 cfm delta triblade, I cannot detect a trace of warmth when feeling the gskill's heatspreaders... This may be why I can run them at this voltage. I'd like to try 3 volts but I'm wondering if I'm pushing my luck and might actually kill them

I remember reading that GSKILL uses only TCCD chips in their memory.. if I'm wrong then I would say they use TCCD for the most part.

chubbyfatazn said:

I remember reading that GSKILL uses only TCCD chips in their memory.. if I'm wrong then I would say they use TCCD for the most part.

Click to expand...

They are not samsung tccd. RedDragonXXX Knows his ram he has several excellent stickys to verify this.

Well sorry then.

Switching these out for the 2-3-2-5 g.skill variety works much better after setting them up according to this g.skill guy's settings

http://www.newegg.com/Product/Product.asp?Item=N82E16820231032

http://www.xtremesystems.org/forums/showthread.php?t=77205&highlight=g-skill

It appears my board is finicky, however, perhaps due to the BIOS... my friend's DFI Expert is running my red g.skills now and is having a much easier time than me. Currently running them at 2-3-2-5 200 MHz and 2.6something volts.

Right now I have the light blue ones running at 212 MHz, 2.62v... and I project I can take them much higher, but I have no need to maximize them since at 212 they are concurrent with the proc being at 2.75 GHz, which is near its max for my cooling, so chances are the proc will max out before the memory. For some reason my Expert only wants to do this @ the 140 7:10 RAM divider. Anything else and it will not dare let me go 300+ FSB. He he... gotta love it.