Danke pc ice.
Hows da lizzard I see ya holding?
Printable View
Danke pc ice.
Hows da lizzard I see ya holding?
lol :p:
hows are those high voltages? whats your max vdimm now, 3.5v? does modding the 3.3v line alone increase the vdimm voltage by the same level you raise the 3.3v line?
2.8 in bios=2.9vdimm
would 2.8 in bios and 3.3v line@3.4v equal in 3.0vdimm?
oh right i forgot your on an cw
Hell Fire maybe I've read over it, but @ what voltage runs that Barton 2505.69Mhz out of the box? :D
I got myself a DangerDen WC kit (Maze4 and NB and GPU block) and so I'd like a new cpu to torture... getting a bit tired of this 1700+ :P
I don't get it, my efficiency seems to be very low, around 95-96%.
If i see others here and see that most have around 98% i'm wondering what i'm doing wrong.
I'm running dual channel, timings are 11-2-2-2.0, and CPU interface is enabled.
My mem is 2x256 Twinmos winbond BH-5.
WTF?
The higher the cpu clock the higher the effiency. That's easy. At least these are my experiences... But maybe the nb is limiting us too.Quote:
Originally posted by Zeus
I don't get it, my efficiency seems to be very low, around 95-96%.
If i see others here and see that most have around 98% i'm wondering what i'm doing wrong.
I'm running dual channel, timings are 11-2-2-2.0, and CPU interface is enabled.
My mem is 2x256 Twinmos winbond BH-5.
WTF?
That's not all, some other parameters influence the bandwidth score asking me.Quote:
Originally posted by SAE
The higher the cpu clock the higher the effiency. That's easy. At least these are my experiences... But maybe the nb is limiting us too.
For instance take a look at page #1 of this thread, and look at the bandwith of Starck, he's got 3523Mbs with only 10,5x225.
That's quite a bit more than mine and mine is at higher clock speeds....why? :(
memory timings?
Nope, 11-2-2-2.0 seems to be the fastest settings for Nforce 2 chipsets.Quote:
Originally posted by saaya
memory timings?
I'm leaning towards the idea that it has something to do with memory quality....:confused:
Hmm. Could be... my efficiency ain't as high as 98% :rolleyes:Quote:
Originally posted by Zeus
I'm leaning towards the idea that it has something to do with memory quality....:confused:
The mem needs to be perfectly stable, you need high cpu Mhz and thight timings. Then you will get 98%. I got it this afternoon.. 3657MB/s @ 10,5 x 233Mhz. Thats 98% if I'm right...
I only get 3120MB/s and I'm running @210MHz, 5-2-3-2.0, that's only a tad more than 92% efficiency. :/
Would 11-2-3-2.0 be more efficient ?
Now there's only one way to find out right? :D 11-2-2-2 would be better.. 11 gives me 35MB/s or so more then 5..so yes, it works.Quote:
Originally posted by Devil's Prophet
I only get 3120MB/s and I'm running @210MHz, 5-2-3-2.0, that's only a tad more than 92% efficiency. :/
Would 11-2-3-2.0 be more efficient ?
My experience is it only works on DC..not SC.. But I'm not 100% about that..
My twinmos 50D doesn't work @ ras to cas 2T, it's too tight for that bugger-kind of ram. :)
Well, my logic tells me that IF 11 would be faster, it would be also more unstable, right ? Like: 5 is stable, but 11 not yet. Is that correct or is 11 more stable because it's a "higher" mem-timing ?
Your max bus is 233x16=3728mbsQuote:
Originally posted by Fewture
The mem needs to be perfectly stable, you need high cpu Mhz and thight timings. Then you will get 98%. I got it this afternoon.. 3657MB/s @ 10,5 x 233Mhz. Thats 98% if I'm right...
Your score is 3657 divided by 3728 (max bus) =0.98095494
Let's just say it's 98% :D
My mem is perfectly stable, i did this mod:
http://www.xtremesystems.org/forums/...threadid=17481
Now i have the feeling that my bios sucks but got to find out what goes wrong, trying to flash it. (n00b)
Some say it is more stable and faster.. But I found that isn't true. It might be more relaxed for the chipset, but not for the memory.. I had to up the voltage a tad to run 11-2-2-2 vs 7-2-2-2. It is faster so yes, it harder on the mem I would think.. :)Quote:
Originally posted by Devil's Prophet
My twinmos 50D doesn't work @ ras to cas 2T, it's too tight for that bugger-kind of ram. :)
Well, my logic tells me that IF 11 would be faster, it would be also more unstable, right ? Like: 5 is stable, but 11 not yet. Is that correct or is 11 more stable because it's a "higher" mem-timing ?
I pulled the heatspreaders off my 3200LL Twinx yesterday.. Winbond CH-5 :rolleyes:
Ik have a possibility to buy Mushkin PC3500 Level2 Dual Pack. It uses BH-5 and my memory does 225 11-2-2-2 3.2V prime stable.. That is a bit dissapointing.. should I dump the Corsair and go for the Mushkin? I was thinking I'd go for it..see what happens.
well.. did the vdd mod (1.84) and the newest bios 3716 does his thing ;)
with GeIL 3500 dual chan .. but this mofo wont let me do 3dmark :/ well gotta try ;)
Hey.
Here's mine at 240 2,0-11-2-2 3,16V dual. Bios 18 and yesterday I 've done another bench I yielded even higher results like 3735/3533MB iirc :D
Update.
Timings 2,0-11-2-2. :D
try the 10 multiplier , this is the highest score you can have with abit ;)
Unfortunately I cannot boot at that one and 10.5 and 11. 11.5 is the one it starts working again :(Quote:
Originally posted by xigfrid
try the 10 multiplier , this is the highest score you can have with abit ;)
Hopefully with my incoming barton I can get back these mults...
@pc ice
Thx :D
Working hard on it. Hope I'll get a good barton matching or even beating those clocks and then I will mod my raddy. If it fails then I will rma it hoping to get a better core ;)
Damn, those are some good bandwidth numbers!
Here's my humble effort:
This is using an Asus A7N8X-X SC motherboard.CPU=Athlon XP 2400+, running at 200*11=2.2Ghz. RAM is a 256MB stick of CRAP Kingston ValueRAM DDR400, made by Mosel Vitalic :mad: . Timings = 2.5/4/4/6. Tried lowering to 2.0, board hung on me. CPU is at 1.675V, RAM @ 2.8V.
Will increasing the RAM voltage to 2.9V help bring the CAS latency down to 2.0?
Finally got some results to show even though they're not that good.
here, this is all i could get with this old 1700 dut3c without the vcore mod since the highest vcore on this board is 1.85v (1.92 for real) and it needs 2v for 2300mhz...
3400 is not bad for 225fsb with 2-3-3-9 :)