Uppdate!
2x1GB Corsair CMX1024-3200C2 v1.2 - Infineon BE-5 @ 2.7v
Validation : http://valid.x86-secret.com/show_oc?id=35949
http://home.online.no/%7Eandr-sei/Fosco/DFI/32m250.JPG
Printable View
Uppdate!
2x1GB Corsair CMX1024-3200C2 v1.2 - Infineon BE-5 @ 2.7v
Validation : http://valid.x86-secret.com/show_oc?id=35949
http://home.online.no/%7Eandr-sei/Fosco/DFI/32m250.JPG
These brackets are really handy, I own a bunch. $15 will actually get you 4 of these (brackets only - no fan just $3.75 each) over at Best Byte. Shipping to California is only about $5.Quote:
Originally Posted by Perc
Uppdate again!
2x1GB Corsair CMX1024-3200C2 v1.2 - Infineon BE-5 @ 2.76v
Timings : 2.5-3-2-6
Validation : http://valid.x86-secret.com/show_oc?id=35952
http://home.online.no/%7Eandr-sei/Fo...I/32m25025.JPG
Quote:
Originally Posted by gundamit
He's waiting on fed ex now... OK memory arrived. it is NOT samsung TCCD, It is not even samsung at all. It is Micron -5B D. Same stuff as the ballistix. He said he would try to OC it and get back to me.
is there anywhere to buy Crucial at lower prices, or get them to resell??? (i work at a computer store, and I'd like to resell these Ballitix!!)
thankx in advance,
~EinZteiN
Regarding the Buffalo Ram, there are several 1GB PC3200 models out there (click for pic below), two of them are both on sale currently, one with Micron, one with Samsung Chips.
The ones i bought in Germany:
DD433-1G/MD with Micron 5B-D
Two Samsungs, not sure but i think those with UCCC are the currently available ones.
DD433-1GSBJ with Samsung UCCC
DD433-1G/SB with Samsung TCCC, NOT TCCD ;)
no longer sold i guess:
DD4333-1G/MC with Micron 5B-c
Oh and i finally got my OCZ's back, running quite well now for theese (see Sig.) .. but i still feel like :slapass: .. > 5 Month to get working Modules that dropped considerably in price since then ...
Nice pics on your post. Bookmarked the CMTLabs page for future reference. Most interesting was the TCCC 1 gig shot. TCCC on 512mb sticks has clocked well in some cases, but I have never seen a 1gig with this IC tested. Apparently this is a pretty rare animal and I only found one Spanish language messsage board that had it listed in someone's sig running at 233 cas3.
Some more searching and I could only find one place in the U.S. with 1gig TCCC sticks. So who wants to take a chance on $118 sticks from Dallas Memory International? Ressellerrating is 7.5.
Here's my attempt with my gimp s754 cpu. Not too bad for a single channel setup, at least IMO. Passes dual channel systems in Everest, for whatever that is worth :)
I can't fully test my ram, since my CPU/memory controller is the limiting factor, I guess. The highest my CPU will stabily overclock too is 9 x 267 = 2403MHz, or 8 x 280 = 2240MHz (which yeilds better benchmarks surprisingly).
http://valid.x86-secret.com/show_oc?id=36066
<a href="http://img46.imageshack.us/my.php?image=280mhz5rn.jpg" target="_blank"><img src="http://img46.imageshack.us/img46/8456/280mhz5rn.th.jpg" border="0" alt="Free Image Hosting at www.ImageShack.us" /></a>
Welcome to Xtreme Diverge !! :toast:
Good results with the Ballistix !!
you can check to see if its your controller that is limiting you OC by test with 2t and seeing if you can take the ram higher. IMO with the Newcastle that it is limiting your memory.
I just got the Buffalo w/ Micron -5B D chips. They did not clock well for me, not even 233 MHz on a DFI Ultra D. I will stick to my 4*512 MB TCCD @ 270 2.5-3-3-7-2T and 4*512 MB UTT CH-5 @ 245 2-2-2-8-2T. They are going back to Newegg, not because they don't clock well, but because they were listed as TCCD and that is not what I received. I may exchange them for Ballistix, but don't really feel the need to.
nice diverge!
did you buy those ballistix in a matched set or individually? i will definitely be getting mine tomorrow so i should have some results within 24 hours.
@ Major, thanks for the tip about trying 2T to see if it is my mem controller. I'll give it a try tomorrow most likely.
@ Sniggle, I bought them individually from newegg last week. So far I like them better then my old muskins :)
Edit: I just tried 2T, and it won't run stable above 280Mhz. I even dropped my cpu multipier down to 6 to take that out of the equation. So, if I don't get any more stability by going to 2T, can I pinpoint it to the mem controller?
I got the same Buffalo sticks and have been playing with them for a bit. Not the cheapo PC4000 I had hoped for, but they seem to be doing better than your sticks. I'll probably return them to NewEgg even though for $100 sticks they are decent performers. If nothing else they got me off the bottom of the list on page one. :woot: 240@2.8v.Quote:
Originally Posted by josanai
http://valid.x86-secret.com/show_oc?id=36149
http://img259.imageshack.us/img259/6...d240x101wx.jpg
your A64 tweaker is fubared
Whoops. Didn't even notice. I guess it doesn't matter since the Buffalo is definitely going back to NewEgg. I just started testing the $200 2gig Patriot set I got from Outpost. Early results are encouraging. I also like the fact the Infineons will do tight or loose timings. 262@2.8v! Finally got my cheapo PC4000 2x1024 set! Yay!Quote:
Originally Posted by Lood007
http://valid.x86-secret.com/show_oc.php?id=36163
http://img357.imageshack.us/img357/6...on262x97ec.jpg
wow! not bad for the price!
I agree, very good results gundamit !
thanks Major :)Quote:
Originally Posted by Major
32m stable @ 2.5-3-2-7 on the asus.. i bet i could get better if i knew jack about a64 tweaker :confused:
well i got my sticks of ballistix today, and so far things are not going well at all.
with my old ram my system was very stable at 2.45ghz with the ram running at 204mhz. now, i can't even prime95 past 1 minute at 2.35ghz with ram at 235mhz 3-4-4-12! i was in windows and able to do a superpi 1M at 255mhz, but nothing was really stable then either. i really hope i didn't get a bad set (although they are not matched sticks, unsure if that really matters).
When testing ram, always use a low CPU multiplier to attempt to rule out the memcontroller as the issue. I and a lot of the other guys here will tell you that the best thing to do it memtest and prime test each stick individually to find each one's max OC.
You may want to try 3-3-3-8 timings and 2.8vdimm. I agree it's not a bad idea to test each individually.Quote:
Originally Posted by sniggle
well i did try a 9x multi with up to 260mhz fsb which was stable enough to bench, but not stable overall. here's a screenshot with various other speeds tested in sandra:
http://funsicle.com/pics/ballistix/245mhz.png
i tried changing the slots and am stress testing now. if all else fails, i will test each stick individually at a 9x multiplier.
edit: so far so good at 240x10 3-4-4-8 1T (on test 3 of prime95, which is a first). i have a feeling that one of the sticks may not have been seated securely, although i couldn't imagine even being able to POST if that was the case. now, if this really is fixed and i can get these babies up to 260mhz+ AND the much touted 78.03 drivers would come out, i'll be in heaven :D
I did some tweaking to my original ram timings and came up with more stability, and more MHz :)
Bumping Max Async Latency from 7ns to 8ns made my 280MHz setup prime95 stable, memtest stable, and even enabled me to get 285MHz prime95 stable. I do get a couple errors in memstest test#5, but only 9 errors over 47 loops, and they always occur at around 2000MB or around 0MB, where there transistion happens i guess. But it is prime stable, overnight... Any ideas what to tweak? :)
http://valid.x86-secret.com/show_oc?id=36279
http://img394.imageshack.us/img394/3...5mhz8mj.th.jpg
Like I said, the only surefire way to test your ram and know its limits 100% is to test each stick individually. Otherwise you are almost shooting in the dark.
Yes, it does take some time, but in the end you know the weakest link and you can set your FSB accordingly.