4x1 GB Crucial Value DDR2-667 @ 400 MHz CL 4-4-4-12, 2,15V
Printable View
4x1 GB Crucial Value DDR2-667 @ 400 MHz CL 4-4-4-12, 2,15V
2 x 1 GB Ballistix PC8000 D9GMH $115:
Boots and runs without error DDR1000 5-5-5-15 1.8v
I run it at DDR667 3-3-3-9 2.0V for 1:1 and voltage for stability.
It "feels" snapier then running DDR800-1000 at higher latencies.
Besides, my 965P-DS3 just doesnt like mem multi of 3's and 4. get cold boots and restarts back to default BIOS :shrug:
For 24/7 use, I have my system running at 450FSB 1:1. I have dropped the 2G Crucial 10th Anniversary ram voltage (which are supposedly handpicked D9GMH) to 1.9v. The system is Orthos stable with 4-4-4-12 timings.
-------------------------------------------
Asus P5B / E6300 @ 450FSB X 7 = 3150MHz @ 1.312v / 2x1G Crucial 10th Anniv. 900Mhz 4-4-4-12 @ 1.9v / Seagate 250G & 320G / X800 XL video / Soundblaster Audigy SE/ Hec AcePower 420w
Corsair 8500
I have reduced my 24/7 voltage from 2.16v to 2.13v (dmm measured).
980 MHz @ 4-4-4-12-3-25-3-9-5-9
475@5,5,5,15 1.9v.
anyone else find it strange the amount of people who bought D9GMH overclocking ram and are barely even pushing them at all. It's like clocking a c2d at like 2.8ghz and saying I've think I've gone far enough.
not really cos im one of them :D Its clear from many results that even 2.15 - 2.2v 24/7 can kill D9Gxx ram, even using aircooling on them. It takes about 12 months, but it happens. More that ~2.2v will just do it quicker. I cant afford to kill them. Besides, with high-bin D9, you dont need to push the volts to run fast with nice timings for eveyday use.
if clocking a core2duo over 2.8 meant it would pretty much die in 12 months or less, you would see a lot of 2.8 or under clocks.
P5B-Dlx when under 24/7 SS: 470MHz (940) 4-4-4-12-20 @ 2.15V 4x512MB GoodRAM Pro PC6400 C5 (D9GMH)
P5B-Dlx aircooled: 400MHz (900)4-4-4-12-20 @ 1.9V 4x512MB GoodRAM Pro PC6400 C5 (D9GMH)
eVGA 680i LT SLI aircooled: (900) 4-3-3-12-20 1T @ 2.2V 4x512MB GoodRAM Pro PC6400 C5 (D9GMH)
[X] 2.1 - 2.19
2.1v for 400mhz @ 4-4-4-10. :)
http://img362.imageshack.us/img362/7299/ddr1200xm3.jpg
Tested later with 2.2V and its rocksolid.
Rbs..
2.35v for 600 MHz 5-5-5-12 24/7
Team Xtreem PC2 5300
4x1 GB DDR2 Crucial Ballistix Tracers PC8500, 1066MHz
BIOS setting = 2.2v
Actual = 2.18~2.20 (via Abit uGuru utility)
I do this to run 4-4-4-12 @ 1013. Otherwise I would loosen the timings, bump up the speed, and drop to 2.1v. I have a 120mm Panaflo fan blowing on them because even at the 2.2 stock voltage the Tracers run a bit warm.
I will never run 24/7 with the Tracers over 2.25v because I've seen far too many D9 modules go bad from over-volting on all types of mobos, whether auxialliary cooling was employed or not.
how long did it take to kill it @ 2.15v? ever go higher than that?
2x Buffalo FireStix DDR2 PC2-8000 @ 4-4-4-12/380MHz/2.2V
2x Crucial Ballastix DDR2-800 @ 4-4-4-12/380MHz/2.2V
Still fine-tuning the machine. :S
4x1 GB DDR2 Crucial Ballistix Tracers PC8500, 1066MHz @ 3-4-3-7 / 400MHz / 2.25v 24/7 Stable
Still playing around with them, won't boot at 1200MHz at all even at 5-5-5-15 and 2.55v (Might be PSU as its at its limits, but doubt it), and haven't tried anything else yet
The previous owner benched them around 2.5v. I bought them, stocked them up for over half a year. Most I ever went was 2.4v just to find max/experiment then to 2.35 for a while. I ended up at 2.2v after a week then after a month or two managed to bring them down to 2.15v.
The mobo temp (sensor around ram area) was always very low even with 92mm on low speed. It just puzzles me the events that led up to it because (while the pc was on all that time, just restarts in between, no shutdowns):
1) Vdub/encoding kept failing with "memory out of bounds error"
2) Orthos blend/large fft passed easy
3) OCCT mem test passed 30min
4) two instance of memtest had no errors for a few hours.
Then I use the PC as normal for one day, shut it down. And one stick is bad and refuse to post at any speed. So sudden.:confused:
I would say I've been running this setup for 4-5 months, but these sticks are very old.. still confused.
yeah when you try to boot at 1.8vdimm they wont do it. it can happen very fast, from errors & rebooting to dead stick. I'd say the 2.5v benching really hurt them, and the 2.35 - 2.4v runs after you got them would have done more damage. If they have warranty from manufacturer, RMA the dead one.
I never tried to boot them at 1.8v yet. There were no errors in any stress testing program except VirtualDub (which I think was a codec problem or something). There was never any auto reboots. I rebooted/restarted myself, this problem only arose when I shutdown completely and tried to turn it on 4-5hours later.
I'm pretty sure bench at 2.5v for Pi 1m isn't that bad. Even after that, they were not in use at all for 6-7months. After that they've been running avg 2.25v for months with not a single hint of an error (again, except that vdub thing which even use to happen on other systems to me)
I am more and more believing its corrupt spd. Just don't know how to reprogram the "dead" one properly. RMA is last resort, I can't be stuffed running in a loop with Team while wasting time and eventually maybe get a crap clocking pair.
currently running 4gb's worth of crucial ballistix aa663 & aa804 on a p35
@ 450 4-4-4-12 2.1v
before that with just the aa663's i was getting 500 4-4-4-10 2.2v
i cant get all 4 sticks to work at 1ghz with 2.2v's at 4-4-4-12 :(
is that normal on a p35? i thought these boards can control 4 sticks just as easily as 2
eh, you lost me there. if you are talking about flashing ram, i'd be in way over my head..
i figured these sticks would be nearly identical to each other, which is kinda why i bought the 804's (not to mention it only cost me 70 bucks :D).
i was thinking heat played a role in it. i have a spare 120mm i was thinking of throwing in. though i'm not sure how to go about mounting it. any suggestions?