There it is! BH-5 is back again!

Quote:

---

Originally Posted by esdee

you should consider that this new release of BH5 is supposed to be made with a smaller proccess-->less voltage needed

---

I agree, but there was no limit with the bh5. People take it to more than 5v.
less voltage should be an advantage if there was no limit.

Test... test... test...

With VX/UTT Intel laid bricks left right and centre before guys like Hipro5/Eva educated the public bout what worked best.
On AMD64 even, there were certain timings that KILLED VX/UTT performance. One of them was Trfc<15.
Also, the errors may have been heat related, or simply using too much vDimm at such a low FSB... that tends to happen with some rigs for one reason or another.

What I'm saying is... we have seen some ABYSMAL results with UTT on some platforms too, and not always on Intel...

This is such a limited sample we have no idea what might be causing it.

Oh yeh... the whole "Intel never had issues with BH5" claim:
Remember, if reports are correct this is 150nm instead of 175nm... this is NOT really even BH5

[EDIT] Mediocre Grammar

today i ordered me 2x512 TwinMOS for my second system.. i got them for ~160$, nice results, looks like they are quite good. I ordered them @ a friend who has go them also @ home and he showed me some benches.. we managed to get on 230 @ 2-3-3-6 @ 2,65v
so quite nice rams for less money

someone test em out with 3.5+ volts please :)

i will test them hard .. , but 3,5v is a bit too much if you ask me for this rams, but will see i will test them than if i got my 2nd system finished, but i will be able to give them 3,1vdimm on my DFI, will try what i get with that.

are there anychance of these 1a4t's showing up on speed premiums?

Quote:

---

Originally Posted by Bla$nA

i will test them hard .. , but 3,5v is a bit too much if you ask me for this rams, but will see i will test them than if i got my 2nd system finished, but i will be able to give them 3,1vdimm on my DFI, will try what i get with that.

---

cmon :D I run my bh5 3.6v for weeks at a time fine.

maybe 3.3v?

Quote:

---

Originally Posted by Shroomalistic

are there anychance of these 1a4t's showing up on speed premiums?

---

can be found on Speed Premium DDR433 which will reach the market soon. Also a few Speed Premium DDR400 with 1A4T are already out ;)

Hi guys!!

Searched a little on the Twinmos homepage :
http://www.twinmos.com/dram/dram_p_sp_ddr400.htm for Speedpremium DDR400
http://www.twinmos.com/dram/dram_p_sp_ddr433.htm for Speedpremium DDR433

in that table is a row 'Dies'
they confirm what many are speculating DDR 400 is CH5 and DDR433 is BH5...

my 2 cents :banana:

Whatever this stuff is..UTT?...new CH5...new BH5 it kicks the snot out of all the BH5 and BH6 Ive tested on NForce2. 2x 512M that is.

Will be interesting to see if "new" BH5 performs any better 1T-wise, the "new" CH5 sure as #### does.

Quote:

---

Originally Posted by SHiRaKaWa

This is the data I've got.
these tests were done using memtest test 5
system:
3.0gh northwood
Msi neo-2 865 motherboard
zalman psu 400w

AA4T was tested on the same system--> 257fsb @ 3.7v 2-2-2-5

this is for 1A4t
@ 2-2-2-5
2.7v --> 210fsb Pass12 = no errors
2.8v --> 220fsb Pass230 = no errors
2.9v --> 225fsb pass5 = no errors
3.0v --> 230fsb pass3 = 2 errors
3.1v --> 230fsb pass2 = 2 errors
3.3v --> 230fsb pass0 = 4 errors
3.4v --> 230fsb pass0 = 225 errors
3.5v --> 230fsb pass0 = 2233 errors
3.6v --> 230fsb pass0 = 22961 errors
-__________________________________-
@ 2-2-2-10
3.0v --> 230fsb pass10 = no errors
3.0v --> 235fsb pass0 = 6 errors
3.1v --> 235fsb pass0 = 16 error
-____________________________________-
@ 2-3-2-10
3.0v --> 235fsb pass12 = no errors
3.0-3.6 --> 240+ fsb errorserrorserrorserrorserrorserrorserrorserrors
errorserrorserrorserrorserrorserrorserrorserrorser rorserrorserrors


Conclusion:
This chip is certainly not BH-5. It hates voltages higher than 3.3v. It works fine with lower voltages but it does not impress. Also, burning in has no effect on this chip.

If you want a chip that works on low voltages, get your self a TCCD.
If you want a chip that works on high voltages, get a AA4T.
Buying this chip is pointless.
*this test is done with an intel setup. results with amd's are not known.

---

Well....
i've got one Corsair 512 mb stick 3500 c2 rev 1.1 with bh-5 (i have removed headspreader) and they also doesn't like voltages higher than 3,3 v.... some like that some like them not... nobody could tell me how to solve this problem.... they are definitly bh-5s but i am not sure, wheter they are better than the new ch-5 or not.
so long and gn8 from Germany
Rondrian

I also experienced that issue with my bh-5 ... 3.55V and they 're fine... 3.6+ and they throw errors like h3ll!!! :mad:

But My twinmos bh-5 (also my khx) did not like 3.4-3.5V first, but after some serious burn in they were addicted to 3.5V :D

So do a lil burn in with them ... say about 24-48hours at max settings you can run stable, but you can also add some mhz. Some errors are okay, thy should decrease with the hours of burning... :D

Let it burn for at least 12hours non-stop, and don't forget to add a fan to the ram ;)

sae... das habe ich schon probiert.... habe mal nen ganzen tag nen burnin gemacht... also wirklich 24h bei 3,6v ohne headspreader mit lüfter.... auf normalen settings funltioniert der super, abser solbald ich mehr als 3,3v gebe macht der mysteriöser weise fehler....

sae... i've tried to burn in the ram 24h but no result.... it still works fine at std. settings but not with vdimm > 3,3v. that's a bit crazy :rolleyes:

Hmm, weird, really weird. Maybe you 've got a bad chip on one or both of the sticks :shrug:

Or the PCB is damaged somehow... :(

I will definitely try to burn in my ram further at 3.6V. It's always been quite effective to turn up Vdimm only a little not to stress the ram too much at once.

So first try to burn it in at 3.35V at stable clocks with 3.3V, then later on, when errors decreased - up the clocks a lil. :up:

It would seem that only the speed Prem comes with BH-5 now...
http://www.xtremesystems.org/forums/...393#post751393

Check it out and let me know what you think...

Alonso question for you

When this is gonna be available ?

Interested specially in PC3500 / DDR433 Speed Premium :)

too bad they didnt show anything at Cebit

I recently saw a Winbond DH-6 chip on a Maxtor HDD. What does that mean? :)

Its just that the numbering is similar. The chip you saw was the HD cache.

w000000000000t

http://www.oc-wear.de/pd1110801656.htm?categoryId=27

But is 1A4T really BH-5? Becasue mine can't pass Prime95 above 220MHz 2-2-2-5! But they are stable in everything else that i'm doing on my computer... I reached 266MHz 2-2-2-5 @ 3.5v single.

Just got 4 modules :) will test

Twinmos
05064
Tmd7608f8e44b
1a4t

ALONSO :

2nd cpuz window shows mismatch - its 2-2-2-5-7 1T in real
DFI nForce4 Ultra-D, Bios 218
3000+ Winchester
TwinMOS DDR400 CL2,5 @ 2,6V/2,66V (Bios/Real) @ 2-2-2-5

First quick results on above mentioned platform:

2x512MB @ FSB250 2-2-2-5-7- 1T @ 3,36V


Watch out for 1A4T (new production started Week7/05)
in product code.
These are brand new so don´t expect them to be alreay in masses everywhere
Arrived in Germany last thursday and first couples of hundreds already sold out to distribution channels.

So 1A4T started as week 7 2005 ?

TEDY, great initial results. :up:

Could you even clock it higher? Did ya test with more voltage, too? :D

lol sae those arent my results :D

this are

Twinmos
05064
Tmd7608f8e44b
1a4t

3.0V 240 2-2-2-8 25loops #5 ok
3.1 and 3.2 doesnt help 250 2-2-2-8 #5 errors
3.3v 250 2-2-2-8 #5 25 loops OK
3.3v 255 2-2-2-8 #5 errors BURN IN here ?
3.4v 260 2-2-2-8 #5 28 errors in 4 pass then i stopped
3.5v 260 2-2-2-8 #5 few errors

What irritates me is that 1A4T runs well 240 on just 3.0v and later drops....

I'm pressuming this stick ( 3 more to test ) is capable doing around 270 with 3.6-3.7v but i'm affraid to put so much juice. Afterall i did burn my KHX3200/256 while pumping them 3.7-3.8 at 270 2-2-2-x...LOL

Any suggestions ??