Meh - I put it to 266 strap and upped the NB and no DLL or CPU errors. But it it BSOD's. So upped the termination voltage and it boots then black screen, no windows etc. :(
Might have another bash with 333 strap... too drunk n work in the morning though >.<
If not then maybe I'll just work on getting 4.8GHZ stable. It goes for a wee bit in prime but nothing spectacullar :(
~Bex
That's rather bizarre because I've been running my G.Skill @ 1088 with only 2.0v. In fact, it can run as fast as 1105 with just 2.0v, and the only aux cooling I have is the side 120mm fan adapted from my LL G70 case to my A77. If I place a fan on the memory it will go as high as 1115 with the same 2.0v. Adding more voltage only seems to produce more memtest86 errors. For testing I initially run memtest 5 for 10 passes then if it passes, I run the entire test overnight. I also run OCCT for an hour, Prime small fft for an hour, then Prime blend at least 12 hours. The memory has never failed once.Quote:
Originally Posted by Grnfinger
I built a system for my brother and he has the exact same G.Skill memory which behaves very similar to mine. He has an E8600 running @4400 with the ram at 1115, 2.0v.
Here's my system:
http://i61.photobucket.com/albums/h5...95smallFFT.jpg
Quote:
Originally Posted by NLight95
i have agree, g-skill ram hates high voltages, the moment you get past 2.25v you start getting more errors. am making progress on my initially posted stability problem with ram. been playing with nb and fsbt voltages. using occt for quicker error detection before i blast the system with prime. and right now they are @ 1080mhz @ 333 strap with 2v actual(higher voltages didnt help but add more heat. i hope am on a hunch here with the fsbt coz it beats having to rma the board. will post back my findings in a day. juss hope they cud 1200.
1200 might really be a stretch, but you can definitely get more out of the G.Skills by placing a 120mm fan to blow air on them. I find that mem coolers like the Corsair Airflow actually impede airflow to the CPU, unless you're using a blow down cooler like a Thermalright XP-120.Quote:
Originally Posted by nama
I forgot to mention that random instability sets in when I set AI Clock Twister to the strong or stronger setting. I keep it at moderate. You take a slight hit in mem performance but gain a lot more stability. And no need to set all the timings to manual (just more settings to remember). I only set what needs changing. If an Auto setting and my manual setting are one and the same (like my base mem timings), then I just leave them at Auto.
BTW, how hard was it to install the 220 Compact? I thought about picking one up but want to see what the upcoming Nehalem pricing is going to look like. I know my 8500 has more to go, it's that I've hit a temp ceiling with my Xig 1283 cooler.
the 220 compact is real easy to instal, can be done in a n evening (more like 2hrs)... i always had no trouble on my previous maximus when it came to memory subsystem. but this one is giving me hell fine tuning the memory subsystem. am learning from it tho. i have a 120cm fan blowing over the ram/nb area so temp wise am safe, maybe i need to reseat the entire stock heat pipe setup.
That's because the G. SKills are using psc chips which will run up to 1100MHz easily on 2.0v-2.1v. They also hate voltages. Anything over 2.1v-2.2v and they turn to crap. Trouble is there are two revisions of the psc chips. Rev A & B. One will do around 1130 2.0v the other requires 2.1v just to get 1066 stable and doesn't clock as high. There's an indeth review on 4GB kits being done over @ Andantech. This info came from the author. All the testing is not complete yet but he spoke highly of those G. SKill & a Patriot kit.Quote:
Originally Posted by NLight95
Hi forum addicts,
About the middle of Mar '08, I built a new system as follows:
Asus Maximus Formula SE
QX 9750 Extreme 3.0 45nm quad (origonal air cooler)
4 gig Corsair CM2X1024-8500C5D
Asus 3870X2
Raptor X 4 74'2 in Raid 0
WD 80 for backup and data
Audigy 2ZS platinum
Liteon DVD
Antec Modular 850 PSU
Housed in a Lian Li 101
The system has been used for gaming....although, like everyone, I thought I might .......
......anyway...never got there yet. I am in aussieland and went to the USA to visit family for 3 months...May to Aug.
I had never had a problem...it was easy to initially setup...flawless and very zippy compared to the last (AGP). Games like Eve Online and Age of Conan it loved.
Yesterday I turned it on for an evenings gaming......and it refused to go past Dram. I fiddled and fiddled, and found that I could start it (occasionally) with only one stick in.
Once I managed to get it past dram and it fired up..I immediately went to the BIOS. The date had reverted to 2002...the raid had changed to IDE etc etc.
Try as I might...including managing to upgrade the bios to the latest ....I cannot get it to boot easily...and certainly when it does boot...it is with only one stick. I have changed the sticks around...and it will start sometimes...with any of the 4...so I figure it's not the memory.
My gut feeling?.........think something has gone in the motherboard.....
Anyone else with any ideas?
Chnage CMOS battery, date loss is usually CMOS battery dying. Although - try increasing teh voltages to the RAM. Corsair seems to need voltages increased on nearly every build I do these days :down:
~Bex
Had all 4 memory sticks checked..........one is gone, gone, gone........
so I'm hoping thats all it is........got BIOS 1207 and working OK at this stage with just 2 gigs
boots up with no worries.....but I worked through the BIOS very slowly
Cheers
I see MIIF is treating you well. I got a dog for a cpu here; max prime stable so far is 4.4Ghz @ 1.440v! :shakes: Anyway, I've trimmed it down to 4.2Ghz 24/7. I'm running high fsb to compensate, 7.5x560, PL9. This is on the second rampage I got from newegg. I've had to rma the other board since the water incident. Both boards have done 600fsb easy. My 2x2GB G.Skills have been the bottleneck so far but still this board amazes me with what it is able to do at relatively low voltages.Quote:
Originally Posted by Grnfinger
Everything is completely stable, 7.5x560, PL9.
Edit: I'm on bios RF 0410.
I'm very pleased with my choice so far.Quote:
Originally Posted by Zucker2k
Seem to be having troubles tho posting past 500FSB , alot of new options to learn.
GTL's have a math calculation so you no longer just set 0.63x ect, its now + or - mv's to adjust wich only complicates.
My E8600 cannot beat my Q9550 in any bench so its going to ebay today or tomorrow,
I see your still have excellent results running Rampage, I hope to aquire one soon:up:
I hope Asus will be out of stock on my RMA on the SE board. Should have got my shipping email Friday ( actually Thursay but I like to add 1 day to wait time ) So Monday I will start my :banana::banana::banana::banana::banana:in and chewin and carry on about the long delay with my 3870X2 and toss in a few remarks about jumping ship in hopes of a upgrade. I doubt it will happen but I will try.
Lol. Goodluck on that. Someone is offering me cherry picked Rampage Extreme for a very attractive price, it's got me very confused since I'll have to get DDR3 ram too. Choices, choices.... :shakes:Quote:
Originally Posted by Grnfinger
I am also pleased with the Maximus II Formula :)
http://img142.imageshack.us/img142/8...ghz135vgs4.jpg
So it seems like the Rampage or the MIIF are really good for FSB.
Thing is, with the Rampage Extreme, you sure won't be bottlenecked by the RAM.
There is almost zero performance increase using DDR3 other than benchmarking. Most current chips cannot utilize the bandwith that DDR2 provides, when the big N is realeased, that will change but for now DDR3 is a waste.n Also if intel releases a 775 chip with a memory controller that could change things aswell.Quote:
Originally Posted by zlojack
I hear you and agree.Quote:
Originally Posted by trt740
I just meant that when you oc the FSB, you get to a point where DDR2 can't go any higher.
How is the 8600 comming along? I have been through 2 of them so far and I'm not overly impressed with it. My E8400 did 4.4 as you well know and I was hoping a 8600 would yeild a 5.0GHz 24/7 system. Well my 2 chips wont.Quote:
Originally Posted by zlojack
If I can get a DDR3 board for an upgrade with Asus I will be a happy boy.
I have alot of ddr2 here that could be sold off for a ddr3 kit.
In reference to the MIIF vs the Maximus X38 regarding ram bottle neck.
GSkill PC8500 2X2GB sticks
Maximus X38 required 2.2volts for 1080MHz and would not go a drop higher without errors.
MIIF requires 2.1volts for 1103MHz and 1133Mhz with 2.2volts and they might go a tad higher with some timing tweaking.
Mushkin XP2-8500
MIIF requires 2.1 volts for 1255MHz:shocked:
and I'm still trying to find the limit, this is running 5-5-5-15 dram static enabled and trans booster to stronger
Well, the first E8600 I had I returned because it clocked like crap and had stuck sensors. The one I have now is a bit better. I could never get 4.4GHz out of the E8400...it wouldn't get stable. I had 4.3 at 1.41v and on this E8600 I've got 4.4 at 1.34v so far, though I'm still tweaking.Quote:
Originally Posted by Grnfinger
I'm curious to try out a new motherboard and see what it can do with this RAM and E8600, but at the same time, I'm not sure if it's worth dumping more money into a motherboard for a very minor difference in performance.
So, between a Maximus II Formula and Rampage Formula, what would the best overclocking choice be for an E8600? My Maximus Formula w/ Rampage Formula BIOS is very difficult to stabalize above 440 FSB, yet the chip is good for 4.4GHz with just 1.275V so I know it has got more in it and I think the motherboard is holding it back.
Both boards would be excellent choicesQuote:
Originally Posted by KRich
On the Maximus x38 my Q9550 would not post past 425FSB on any multi.
On the MIIF it will not only post but is benchable up to 501FSB on any multi.
I dont own the Rampage X48 but Zucker2k does and he also owns the E8600
on the 7.5 multi he runs a nice 560FSB
Thanks buddy :up:Quote:
Originally Posted by Grnfinger
I guess then that overall, both would not hold the chip back, but the Rampage Formula would probably allow the maximum FSB using a lower multiplier.
The Maximus II Formula has got some nifty new stuff though and it looks like the BIOS has got more features too.
My PC hates me again. *sigh*
5GHz no more CPU ERROR on LCD but sometimes boots, other times not. 4.8GHz hates me as well now so on 4.6GHz atm but even that's went a bit funny..... :(
~Bex
Whats your GTL's at?Quote:
Originally Posted by RCG_Bex
I had 4.1GHz pass linpack 10 loop test twice then it would fail after that.
Adjusted GTL's only and it passing 20 loops 3 times, going to prime it now
Unless you live in an incredibly cold room and can maintain all your temps close to 10C, it'll be next to impossible to run that chip that high 24/7. 4.5Ghz is doable, but even that depends a ot on cooling.Quote:
Originally Posted by RCG_Bex
Hey Zucker2k, what was the max FSB you could run your Q9450 on the Rampage Formula? Still can't get 500FSB stable but this Maximus II is an early production board with the A2 chipset. I'll probably try the E8400 in it just to see where it tops out. I'm guess 550FSB.