PDA

View Full Version : Geil One Winbond!!! Really Nice Sticks


Ozan Baseski
01-31-2006, 05:30 PM
Today I had a chance to test my new Geil One Winbond sticks in my home pc. At first glance I found out that it exceeded my expectations. Really sweet ram :toast:

Test System:

Dfi Sli-Dr Expert
A64 FX57-Vapochill LS
2x512 MB Geil One Winbond

273mhz super pi 1m
http://img485.imageshack.us/img485/2354/273mz1m8mv.jpg

274mhz super pi 1m
http://img71.imageshack.us/img71/4457/274mz1m9qv.jpg

275 mhz super pi 1m
http://img485.imageshack.us/img485/790/275mz1m0bk.jpg

Vincentvega18
01-31-2006, 05:41 PM
very nice, what kind of volts?

Ozan Baseski
01-31-2006, 05:51 PM
It's around 3.75-3.85 volts

eLias
01-31-2006, 06:01 PM
Probably they will die soon just like all VX and RedLine :(

Major
01-31-2006, 06:25 PM
I looked to buy these sticks a couple of months ago and couldn't find them anywhere ?

Pyr0
01-31-2006, 06:27 PM
geil have discontinued them :(
http://www.geil.com.tw/portal/news120805.php

[cTx]Philosophy
01-31-2006, 06:44 PM
I would love for winbond bh-5 or ch-5 to come out on 1G sticks that would own definately..

[XC] moddolicous
01-31-2006, 06:59 PM
Newegg use to have them, but a quick check doesnt show them. These are the same ones Lardarse/Shmamino uses for his 3dmark runs, right?

s7e9h3n
01-31-2006, 10:56 PM
I would love for winbond bh-5 or ch-5 to come out on 1G sticks that would own definately..
It's not possible to have those IC's on 1gb sticks. I don't remember why, but that's the same reason we don't see 1gb TCCD sticks.....

Ozan Baseski
01-31-2006, 11:37 PM
Probably they will die soon just like all VX and RedLine :(
I hope it won't soon and of course I don't use those voltages for daily use. I use them at 250 mhz with 3.3 volts and active cooling all the time

Coroner Kyle
01-31-2006, 11:41 PM
Why did your SPi times increase while you increased speed? Did you change memory settings?

Ozan Baseski
01-31-2006, 11:51 PM
I increased memory speed but decreased the cpu speed if you look at the SS's more carefully bc my FX57 does not stay stable above 3550 mhz on this stock Vapochill cooling.:)

Coroner Kyle
02-01-2006, 12:00 AM
Ok, sorry I must have missed that my eyes are starting to get blurry. Stock LS is weak. :)

largon
02-01-2006, 12:07 AM
Ozan Baseski,
Tried dropping tRTW and tWTR to 1?


s7e9h3n,
The reason is simply that Winbond W942508BH/CH and Samsung K4H560838F-TC__ are too low density.
(chip density 32x8 means 256Mbit/32MB per chip).

Since normal DIMMs can hold only 16 chips (16*32MB=512MB)...

Well, actually a 1GB dimms could be built using these chips but that would require stacking ICs and the dimm would be equally straining to the memctrl as two 512MB DIMMs.

s7e9h3n
02-01-2006, 12:10 AM
I increased memory speed but decreased the cpu speed if you look at the SS's more carefully bc my FX57 does not stay stable above 3550 mhz on this stock Vapochill cooling.:)
You should be able to get a better time @ 273 though - for comparison, here's 13x270:

http://www.cryo-laboratory.com/upload/userfiles/s7e9h3n/TEAM%20BH5/13x2701m-.jpg

Ozan Baseski
02-01-2006, 12:15 AM
@Largon - I didn't try them yet but will give it a go and see how they would do with those timings.

@s7e9h3n - I didn't fully tweak for a super pi time, rather I wanted to see the 1m stability of the chips. You are definitely right that I would get a better result than 23.50 sec.

s7e9h3n
02-01-2006, 12:33 AM
s7e9h3n,
The reason is simply that Winbond W942508BH/CH and Samsung K4H560838F-TC__ are too low density.
(chip density 32x8 means 256Mbit/32MB per chip).

Since normal DIMMs can hold only 16 chips (16*32MB=512MB)...

Well, actually a 1GB dimms could be built using these chips but that would require stacking ICs and the dimm would be equally straining to the memctrl as two 512MB DIMMs.
Yeah, that's what I meant to say...lol...:toast:

uOpt
02-01-2006, 05:51 AM
Are you the guy who just bought mine on ebay?

The timings and frequencies reached look familiar.

Ozan Baseski
02-05-2006, 03:47 AM
Yesterday I tried them with tighter timings and here are the results. That's the best it can do for now I think. I did the test with 3.88 volts. These ram are more like old bh5 bc my previous geil one winbond sticks don't boot above 3.7 volts no matter what I do and they were not this much lucky of course.

http://img342.imageshack.us/img342/4711/270mz1magg8lk.jpg

yotomeczek
02-05-2006, 03:50 AM
Today I had a chance to test my new Geil One Winbond sticks in my home pc. At first glance I found out that it exceeded my expectations. Really sweet ram :toast:

Test System:

Dfi Sli-Dr Expert
A64 FX57-Vapochill LS
2x512 MB Geil One Winbond

273mhz super pi 1m

274mhz super pi 1m

275 mhz super pi 1m


Nice but show ua 16 or 32M ....
1M is not big challange for memory

s7e9h3n
02-05-2006, 03:57 AM
Yesterday I tried them with tighter timings and here are the results. That's the best it can do for now I think. I did the test with 3.88 volts. These ram are more like old bh5 bc my previous geil one winbond sticks don't boot above 3.7 volts no matter what I do and they were not this much lucky of course.

http://img342.imageshack.us/img342/4711/270mz1magg8lk.jpg
Hold on a second.....you say you're using an Expert? I ask this because it's not possible to run @ cl1.5 on the Expert at that speed. The Expert runs a TRUE cl1.5 and it's barely even possible to boot @ 230Mhz at that latency. The older Ultra/Sli/Sli-DR's were able to run that CL1.5 since that was in fact a slower setting than CL2. If your run was on an Ultra-D, your timings would in fact be slower than running cl2......Here - this is on an Expert:

http://img232.imageshack.us/img232/1149/13x2701m7lc.jpg

Ozan Baseski
02-08-2006, 01:23 AM
Yes my friend. I was running these sticks at 1.5-5-2-2 on the expert. At least it's what cpu-z is saying to me.

s7e9h3n
02-08-2006, 02:51 AM
Yes my friend. I was running these sticks at 1.5-5-2-2 on the expert. At least it's what cpu-z is saying to me.
Hmmm...could you get a cpu-z validation for me? Your 1m time should have destroyed mine @ the same speed :confused:

NickS
02-08-2006, 12:54 PM
Nice result Ozan. That Geil One really kicks :D.

M3kk
02-10-2006, 10:51 AM
Ultra-D AD0 , VDIM 3.66V
http://valid.x86-secret.com/show_oc?id=70733

http://img496.imageshack.us/img496/3324/gonespi3226515225tw25rn.jpg (http://imageshack.us)

http://img496.imageshack.us/img496/6995/gone265a64tw5zp.jpg (http://imageshack.us)

irev210
02-11-2006, 05:49 PM
i have the Geil One winbond's

I cant pump more than 3.52...

still able to pump 2-2-2-5 @255

YMMV

M3kk
02-11-2006, 11:37 PM
Try to put mem`s t in the orange slots ..

s7e9h3n
02-12-2006, 02:44 AM
Try to put mem`s t in the orange slots ..
Ummm...don't try that if you plan to run the sticks >3.2V......

krampak
02-12-2006, 02:49 AM
uhm..... why? I was running 2x256MB levelII 338WF at 3.85V yesterday on orange slots... should I use the yellow ones? It doesn't oc as good as the orange slots! (i'm using a DR-Expert).

M3kk
02-12-2006, 02:53 AM
Ummm...don't try that if you plan to run the sticks >3.2V......

Well.. by me they are more stable in orange slots.. here they do 265mhz 1.5-2-2-5 memtest , and SPI32MB stable.. with 3.66vDIM

s7e9h3n
02-12-2006, 03:08 AM
uhm..... why? I was running 2x256MB levelII 338WF at 3.85V yesterday on orange slots... should I use the yellow ones? It doesn't oc as good as the orange slots! (i'm using a DR-Expert).
Oops...I forgot we're in a thread where the Expert is the board at hand :p: Yes, on the Expert, the sticks should be in the orange slots...On the older DFI NF4 boards, LL/high voltage sticks should be in the yellow. Sorry about that :slap:

Ozan Baseski
02-13-2006, 01:02 AM
@s7e9h3n - No matter what I do I can't get a better score than 23.625 seconds when I use 1.5-0-2-2 timings as you suggested in the tweaker. At 270x13 I can get 23.625 but not below 23.500.

s7e9h3n
02-13-2006, 03:18 AM
@s7e9h3n - No matter what I do I can't get a better score than 23.625 seconds when I use 1.5-0-2-2 timings as you suggested in the tweaker. At 270x13 I can get 23.625 but not below 23.500.
Hmmm...I'm not quite sure what you're asking/saying here :p: 1.5-0-2-2 are impossible timings....I think you may have put them in the wrong order :) Here's another run @ 13x270 with different timings (I actually can run this speed much tighter, but I used these settings in order to just to pass on the first try )

http://www.cryo-laboratory.com/upload/userfiles/s7e9h3n/TEAM%20BH5/13x2701m-.jpg

Ozan Baseski
02-13-2006, 04:57 AM
I'll try your suggested timings. And when I wrote 1.5-0-2-2., 0 value is the value for tras. I can write 1.5-2-2-0 also if you like :p:

s7e9h3n
02-13-2006, 10:29 AM
I'll try your suggested timings. And when I wrote 1.5-0-2-2., 0 value is the value for tras. I can write 1.5-2-2-0 also if you like :p:
OK, I get it now. :D ...let us know your results:toast:

mnewxcv
02-13-2006, 10:34 AM
so what does it do at 2.9v? i have some winbond chips and my board maxes at 2.9v...wanna know what do expect. (runs 230 2-3-2-5 @ 2.6v)

Ozan Baseski
02-14-2006, 01:00 AM
You can't get much with giving only 2.9 volts. These chips show their real performanca above 3.0 volts in fact.

irev210
02-15-2006, 06:49 PM
You can't get much with giving only 2.9 volts. These chips show their real performanca above 3.0 volts in fact.


depends on your memory, mine do pretty well with low volts.


Though the best power/performance ratio is probably at 3.17vdimm


I want 2x1gb dimms. I paid 200 bucks for these geils... i dunno why i cant do any higher than 2-2-2-5 @ 255.

gr.