Shouldn't be a problem, but you would get higher clocks with 3-2-2 and would lose virtually 0 performance.
Printable View
Shouldn't be a problem, but you would get higher clocks with 3-2-2 and would lose virtually 0 performance.
even the 3200eb should be able to do that without too much volts (2.8v)Quote:
Originally posted by RLiu818
Anyone know if the OCZ EB PC3700 can run stably at 2.5-2-2-10 at 200-210 range?
it should without any problem..
*edit didnt see the above posts :stick:
With my Mushkin level one 1x512 at 2.6v, how can I compare the difference of running 218fsb at 2-3-2-11 and 200 at 2-2-2-11?Quote:
Originally posted by Geforce4ti4200
eshbash has tested it out. the other timings matter very little. going from ras-cas delay of 2 to 3 costs 25fsb in performance. all the other timings combined only cost 5fsb. thus 2-2-2-5 at 220fsb was equal to 2.5-3-3-6 at 250fsb. not bad. 3-4-4-8 would need like 280fsb to be equal.
I tried the Sisoft memory bandwidth benchie, but I don't think thats a good way since its bandwidth, not the timings. The 218 did better there. Also did Super Pi for the heck of it, but 200 gained 2 seconds, since I'm at 9.5 multi atm so 1900mhz vs 2071mhz would be the winner over the timings there I guess.
Comparing fps's from a checkpoint in farcry? Any benchie I should do?
The fsb's in this thread are.........are.... :slobber:
EDIT: nvm, pretty close, but reboot after 5 mins of farcry at 2-2-2 :(
It could still see an improvement at 198 or something, but I cant change the vdimm on my SN45G, and anything under 200fsb on AUTO sets to 2.5v so its a no go. My hope lies in the ocz vdimm thing :)
:toast:
Not sure what your getting at, but super pi will be faster at 2-2-2 lower fsb, than 2-3-2-11 higher FSB. Just the way Super pi is.
Anyway, either your chip or board is causing your problems, My 3700EB did 220Mhz 2.5-2-2-11 in my NF7-S no problem.
Just got my new system fired up, and was able to make it thru 2 passes of memtest at 250 FSB at 10-2-3-3, 2.8v. Installed the OS and tried P95, but it failed right away. CPU was at 2GHz and HTT at 4x, so it wasn't that. Set the mem/CPU to stock and now P95's been going about 30 minutes.
I'll have to play around with it some more, and plan on getting the DDR voltage booster when released, but so far I'm not getting the results I had hoped for. I was thinking 250 MINIMUM at x-2-3-3 and 2.8v.
218 2-3-2-5/ 218 may be a tad quicker than 2-2-2-5/6 200.. you gotta play around a test stuff.
My PC3700 EBs 2x512 does 255 3-3-2-8 1:1 @ 2.8v on my stock 8KDA3+.
while my Mushy PC 3500 2x512 BH-5 does only 215 2-2-2 1:1 @ 2.8v
what?, but your comparing diff timingsQuote:
Originally posted by geneh_00
My PC3700 EBs 2x512 does 255 3-3-2-8 1:1 @ 2.8v on my stock 8KDA3+.
while my Mushy PC 3500 2x512 BH-5 does only 215 2-2-2 1:1 @ 2.8v
your bh5 could use some more voltage ;)
Did some more testing since last night. I was able to get P95 to run for 3 hours before crashing instead of 2 minutes at 250 by blowing a fan on the board (it's not in the case right now, it's on the workbench). I was planning on getting the OCZ DDR Voltage Booster when it becomes available, but now I'm wondering if heat seems to be a factor at 2.8v, isn't 3.0v just going to make it worse? Without the fan I can only touch the memory for a fraction of a second, it's that hot.
I set it down to 245, 10-3-2-3, 2.8v, and P95 has been running over 12 hours now. So my limit is somewhere between 246-249, and I'm not sure if it's hitting a ceiling for 2.8v, or if it's getting too hot and more voltage won't matter. I'm going to try 250 again but at 2.7v, and see what happens.
Are you running it with thea heatspreader attatched, or have you removed it?
hmm just bougth 2 x 256mb sticks.. not to impressed I´ll say..
does 252mhz (7-2-2-3 at 3.08V) and 273mhz (8-3-2-3 at 3.6v+)
I´ll try some higher volt when running tight timings but only 273mhz at 3.6V with relatively loose timings does not impress me at all.. But hey I got all the weekend to play with this stuff so maybe i´ll improve..
oh its pc3700 btw..:)
im gonna stop all together of doing prime95
the other day i was running my 3700EB (2x256) at 266, 3-2-3-7 at 3,45 with prime95 for like 4 hours no errors then switched off everything
next day went to prime not changing a thing and after 2 minutes ... error ... I was like "WTF"
... so i went along playing americas army for like 2-3 hours with that settings without a single "critical error"
... and yeah i can run 3d benchs at over 270 so i guess ill just use prime for burnin :p
Heatspreaders are attached. Dropping voltage didn't help at all, so I stopped playing around with it for now while I get an idea of what my FX-53 limit will be on water. Once that's done I'll go back and figure out where in the 245-249 range my max is going to be. I guess 245 isn't really too bad, maybe the voltage booster will do some good. Won't hurt to try it I guess.Quote:
Originally posted by gouda96
Are you running it with thea heatspreader attatched, or have you removed it?
thats simply sick
If they are hot remove the heatspreader. In my experience heatspreaders have insulated the chips more than reduced heat. Then you could cut individual ramsinks out of Cu or Al and that would probably keep them pretty damn cool.
Did you know that OCZ PC3200/PC3500/PC3700 EB's are about all the same? It whould cost way more to OCZ to put all different chips and PCB in them. The difference is just marketing, and sometimes, some RAM's fail, and go down a grade. But if you take all those 3, you should have about the same overclock.Quote:
Originally posted by saaya
even the 3200eb should be able to do that without too much volts (2.8v)
So you should not have problem to do a speed/timing w/ PC3200EB that PC3700EB....
Quote:
Originally posted by skate2snow
Did you know that OCZ PC3200/PC3500/PC3700 EB's are about all the same?
So you should not have problem to do a speed/timing w/ PC3200EB that PC3700EB....
they are not the same and they do not overclock the same.. Allthough the difference isnt that big its there.. :)
What do you think, that they will make 3 different RAM's, and rate them differently? The only difference is the warranty, some are rated to do higher, and some to do lower, but at the end they are about the same....
I know that for a fact.. perhaps same chips but the 3700 are the top of the line ones´ aimed for a certain timing and Mhz testet on a certain number of motherboards
And its like saying that OCZ is only taking the top quality Micron 5ns chips..... How do you think they know which are bette before they buy from Micron?
And they wont test the chip like that. They will just test if its able to do the speed/timings they whant, and if it pass, they are PC3700EB, if not, they degrade to PC3500 EB, if it does not pass, they degrade to PC3200EB. And an OCZ rep said that its rare that the chips fail to PC3700 tests. The difference is that they are tested differently.....
You see thats the marketing part ;)Quote:
Originally posted by skate2snow
And its like saying that OCZ is only taking the top quality Micron 5ns chips..... How do you think they know which are bette before they buy from Micron?
I thats what I believe I just said :)Quote:
Originally posted by skate2snow
And they wont test the chip like that. They will just test if its able to do the speed/timings they whant, and if it pass, they are PC3700EB, if not, they degrade to PC3500 EB, if it does not pass, they degrade to PC3200EB. [/B]
But, Not much fails. They just take SOME not ALL at PC3700, and some at PC3500, some at PC3200 and if ones fail, they degrade....
that "some" you are mentioning are not randomly choosen.. but yes cant argue with the rest..Quote:
Originally posted by skate2snow
But, Not much fails. They just take SOME not ALL at PC3700, and some at PC3500, some at PC3200 and if ones fail, they degrade....
just got my 512MB PC-3700 stick yesterday and did some benchmarking straight out of the box.
VDIMM was set to 2.7V(max from bios) but this K8NS PRO overvolts a little so it´s actually 2.75V.
265MHz 3-3-2-8, 2T, benchmarked pifast and superpi 1M
255MHz 3-3-2-8, 1T, benchmarked pifast and superpi 1M
250MHz 3-3-2-8, 2T, superpi 8M
I´ll have to say that I am really impressed what this memory can do, even with 2.75V! I didin´t try running prime95 for stability, just tried some limits first... oh, I also didn´t try any lower latencies but I´ll try them later..
I´ll need to do some burn-in for couple of days and then see what this baby can do... though have to fix vdimm mod today to provide ~3V...
Couldn´t be more happy with this memory....:D
btw. performance is great even with those 3-3-2-8 latencies compared to my BH5 2-2-2-10...
:banana:
has anyone used the A64 tweaker with EB memory? I was just figuring which would be the fastest settings?