TEAM Xtreem BH-5 Review
_________________________________________________

TEAM Memory introduced their new line of memory called TEAM Xtreem series of memory. These modules are based on a couple of ICs. They have modules based on Samsung TCCD, Samsung UCCC, BH-5 and Infineon BE-5.

This TEAM Xtreem BH-5 kit uses Brain Power PCB. These B6U808 PCB has better trace lengths and signal properties than JEDEC's reference PCB which translate into better overclockability. Another plus for these modules!

To fully take advantage of BH-5 memory. A higher Vdimm needs to be supplied to the memory. Most motherboards won't be able to supply enough Vdimm to take advantage of modules based on BH-5 ICs. So far only DFI and certain Abit motherboards gives the options of running Vdimm of more than 3.2v. For other users, the other way to get more Vdimm is through the use of OCZ's DDR Booster. Please check OCZ's motherboard compatible list and check that your motherboard is compatible with the booster before attempting to use it on your motherboard. Doing so might kill certain hardware in your system. Please make sure that your powersupply unit have a -5v line before using the DDR Booster. If you do not have the -5v line the booster's display will not work and the booster's power circuit may not function properly too.


Web Link : TEAM Xtreem BH-5
Price : AUD179 = RM540


Information from TEAM
The 'Xtreem' series is, as the name suggests, a range of extreme-performance RAM modules designed for extreme players, especially those who enjoy high-frequency or 3D games.

The Xtreem BH-5 is built using the Winbond BH-5 UTT IC Chip, a legendary special generation chip manufactured under 0.175 micron processing.

These memory modules achieve superb performances by exploiting its special characteristic of being able to withstand high voltage operation (over 3.0V on a 2.6V standard)

Due to its continuing high demand and popularity, Team Group is bringing this sought-after chip back to the market, especially for you overclockers and hardcore gamers out there with the need for speed.

The 'Xtreem BH-5' itself is a fine masterpiece from Team Group, resulting from our painstaking work to adjust the SPD in order to operate at 400 MHZ, CL 2-2-2-5 in Dual Channel mode without the need to tweak any of your BIOS settings.( These latencies are significantly faster than existing modules in the DDR memory market available today).

Each Xtreem module (as with any TeamGroup ram module) is lifetime - and satisfaction - guaranteed.
Spec :
Model / Speed : PC3200 DDR400MHz
Capacity : 512MBx2 (dual channel)
CL/ Speed : 2-2-2-5


Modules still in packaging


Sticker Pic


Close up on chips


Picture of the kit


System Setup

Test Setup :

AMD Athlon FX-57
DFI NF4 Ultra-D (Bios 704-2BTA)
Antec 550W TruePower2.0
XFX 6600GT
2 x WD Raptor 36Gb Raid 0
Corsair Cool WaterCooling Kit

Since new DFI NF4 bioses offers new divider options that allows users to run your memory above FSB e.g. 4:5 that allows ram to be clocked at 250Mhz and FSB at 200Mhz. Tests will run at DDR400 DDR433 DDR466 and DDR500 with the CPU at the same speed and highest overclocked memory speed.

DDR400/433/466/500 are only available through bios. CPU-Z, Memtest, A64 Tweaker and other software are unable to detect those settings. A screenshot of the startup splash screen showing DDR400/433/466/500 will be provided.

This post has been edited by babyelf: Oct 14 2005, 04:04 AM

This post has been edited by babyelf: Oct 14 2005, 04:00 AM

This post has been edited by babyelf: Oct 14 2005, 04:01 AM

is it 512 a stick ?

how much are a pair ?

and btw , are u sure they are teh xtreem version ?

coz they dont hav teh heat spreaders like tha one at their site

This post has been edited by zx7177: Oct 1 2005, 03:07 PM

duno about Malaysia but here it's about the price of OCZ's Value VX

and yes they are 512Mb x 2

This post has been edited by babyelf: Oct 1 2005, 03:12 PM

look @ the sticker

i think the new ones don't come with heatspreaders anymore

still at 2.8v

omg still 2.8vdimm...
this ram gonna fly if applied with 3.2vdimm

Evogenix

err.. n00b asking here. what so special bout tat?

@babyelf
how much is it???

bah.....my bh-6 can past at 225mhz on 2.9v(windows detect at 2.86v)

then how about the timing ??

DDR444 @ 2-2-2-5 @ 2.8V is quite good already

answered a few post before about the same as OCZ Value VX

alpha timings?

This post has been edited by babyelf: Oct 2 2005, 12:47 AM

bah.....my bh-5 can past at 230mhz on 2.9v(windows detect at duno wat voltage yet)

this bh5 look similar (physical) with my gskill GH. unfortunately utt bh5 have big issue with abit nf7 v2.0, if not, i also wonder which is better between this baby and GH.

afaik UTT have issues with nf2 systems..

so it should have the same issues as the GH..

for NF2 old school bh-5 still performs the best

bumped voltage up to 3.1v

241 is max on 3.1v





now going for 3.2v

@babyelf
thnx for the info

regards,
aaron

no problem aaron

currently burning in 3.3v 250mhz

started out with 618-3 bios.. was lazy to change the bios first.. but then have to change it now cause i'm pretty sure it's the bios holding me back.. so flashed to 704-2BTA and it's running speeds that weren't possible with 618-3 bios..

250mhz 3.3v gave some errors before.. but i left it running for about 3-4 hours now.. and the errors are gone

3.3v 250Mhz...

errors i got before is gone now

not a bad burning in session for today

if u dont mind can u share ur bios setting on dfi nf4

thx

it's on the first post

thing i changed is twtr is 2 instead of 1 now..

twtr of 1 topped out at about 242-245mhz.. had to loosen it to 2 to get higher

Those are really nice babyelf. You probably have as much good ram as anyone. I hope those continue to improve and someone in US starts selling them!

Thanks Coop..

yea hopefully.. they are famous in taiwan

they just launched a new website as well

www.teamgroup.com.tw

burned in the ram a bit today

256 gave me errors before.. 257 gave me more error

but this is 256 now





This post has been edited by babyelf: Oct 5 2005, 01:48 AM

w00t still a 4mhz increase per V now..

3.5V

w00t

3.6v

why not just let it burn at 3.5V?

nice nice nice........ !!!!

a better altenative to twinmos sp............ but hopefully cheaper than ocz goldVX.

it is.....

i usually let it run until the errors go away at a certain speed and voltage..

here is 3.7v.. it's still going

this is about the price of Value VX.. Gold VX is much more expensive





This post has been edited by babyelf: Oct 5 2005, 03:50 PM

10 minutes into test 8 now

32M

nice one bro...my ocz vx value could never reach that speed with that setting...got 265@2225@3.6~3.7 but the setting not as tight as yours...i have to loosen up the setting a bit...this TEAM stuff seems to be a very good buy so, i guess another ram find a place in ur collection item list huh
any bulker???

This post has been edited by remysix: Oct 5 2005, 09:28 PM

lol yea.. added to my collection.. when i finish testing might take the UCCC out to play next

looks like a sweet pair of rams

regards,
aaron

really sweet

2.59v 200mhz 2-2-2-5

bro...just curious...can u get the same result wif diff proc...last time, using my 3200+ E3...i can hit 265@2225 memtest stable...but know when i used this 3000+ E6...only 260...not even 261 without getting errors in memtest...i think this is one of the cons of having mem controller build-in the proc...changing the proc might give u diff result...

This post has been edited by remysix: Oct 7 2005, 06:25 AM

yea.. memory controller plays a very big role on ram clocking..

and another thing that will affect ram clocking is temps.. lower temps will clock ram higher.. cooler memory controller will go really high

looking forward to my phase change unit soon.. should be sweet

well to prove that memory controller does differ

my FX is clocking the ram at 266mhz now

good for you...last time u pair with which proc???
that 266 at what voltage bro...

at 3.7

but not stable.. only memtest stable tho

reminds me of my first kit of 1Gb TMSP UTTs that could do 263Mhz.. Miss those times....

This post has been edited by neuvas: Oct 12 2005, 06:44 AM

r u telling tat ur RAm speed is 266Mhz/ ddr532 when paired wif FX?? n00b in wondering here

I believe that is what he means

since he can do memtest stable, some tweaking to the memory timings will make things more stable. these extra memory timings are actually related to the memory controller, not just the memory module.

and the worst part is...whenever u changed ur proc...u almost need to do it all over again...and the previous setting with previous proc might not valid anymore, thus achieving the same result MAY not be possible...and that's just plain suck

well it sucks when you don't have the processor that clocked it higher before

if you have a new processor that clocks it better u won't be complaining imo

babyelf when u gonna test UCCC chip? seems like a high potential OC ram

there are 2 better chips than UCCC

there's ballistix (but micron won't sell their chips to other manuf)

and there's the new OCZ EB PC4000 kit

i'm thinking of about 280+ 3-4-4-8 for the TEAM UCCC

hehehe...i guess u'd know why i'm complaining...

lolz........

updated first, second and third post with benchies.. one of them will have to have a rerun.. must've left something running and made the score lower..

note that i'm using new dividers that runs ram at a higher speed

DDR400/433/466/500

hmm...how does this BH5 performs on Intel? any info? dis is NEW BH5 rite...so theoretically it performs better only on AMD compared to Intel...

i'll see what i can find out

any updates on this?

told u on msn

it's tested to run stock.. but overclocking wise i'm not sure.. i haven't had an intel rig for more than 7 years lol.. last one P3

OPB testing some TEAM BH5

http://xtremesystems.org/forums/showthread.php?t=82109

TEAM BH5 owners

mind posting some preliminary results here?

my preliminary result with Team BH-5

255Mhz@3.6vdimm X 11 =2.8Ghz


Attached thumbnail(s)

no pics but this is what i get, juz started to play w/ 'em yesterday...

2.6V > 200MHz

2.7V > 209Hz

2.8V > 217MHz

2.9V > 223MHz

still burnin...

any tip guys? 223MHz w/ 2.9V is not so impressive...

Below is my DRAM settings:

Dram Frequency Set (MHz) : 166
Command Per Clock (CPC) : Enabled
CAS Latency Control (Tcl) : 2.0
RAS# to CAS# Delay (Trcd) : 02
Min RAS# active time (Tras) : 05
Row Precharge Time (Trp) : 02
Row Cycle Time (Trc) : 08
Row Refresh cyc time (Trfc) : 17
Row to Row delay (Trrd) : 03
Write Recovery Time (Twr) : 02
Write to Read delay (twtr) : 02
Read to Write delay (Trwt) : 02
Refresh Period (Tref) : 166MHz 1.95us
Write CAS Latency (Twcl) : 01 Bus Clock
DRAM Bank Interleave : N/A

DQS Skew Control : Auto
DQS Skew Value : 0
DRAM Drive Strength : Normal
DRAM Data Strength : 50% reduction
Max Async Latency : 7.0 Nano Seconds
DRAM Response Time : N/A
Read Preamble Time : 5.0 Nano Seconds
IdleCycle Limit : 16 Cycles
Dynamic Counter : Disable
R/W Queue Bypass : 16x
Bypass Max : 07x
32 Byte Granularity : Disable


This post has been edited by ah_khoo: Dec 10 2005, 10:09 AM

Nice. Got them running at 2.9v 225MHz more or less, and 230MHz gave thousands of error.... Tested that one stick actually can't keep up with the other, resulting in 240MHz 3.1v MemTest free while the bad gave also thousands of error

225MHz gimme instant reboot, 223MHz ok for bout 20Mins of blend test in go sp2004, memtest of windows version gimme instant error, even @ 200MHz...

will try again tonite. pls share d result guys...

Khoo. Try this settings....

Dram Frequency Set (MHz) : 200MHz
Command Per Clock (CPC) : Enabled
CAS Latency Control (Tcl) : 2.0
RAS# to CAS# Delay (Trcd) : 02
Min RAS# active time (Tras) : 05
Row Precharge Time (Trp) : 02
Row Cycle Time (Trc) : 07
Row Refresh cyc time (Trfc) : 12
Row to Row delay (Trrd) : 02
Write Recovery Time (Twr) : 02
Write to Read delay (twtr) : 01
Read to Write delay (Trwt) : 01
Refresh Period (Tref) : Auto
Write CAS Latency (Twcl) : 01 Bus Clock
DRAM Bank Interleave : Enable

DQS Skew Control : Auto
DQS Skew Value : 0
DRAM Drive Strength : Normal4 (8) --> Normal3 (6)
DRAM Data Strength : Level3 --> Level2 --> Level1
Max Async Latency : 7.0 Nano Seconds
DRAM Response Time : Normal
Read Preamble Time : 5.0 Nano Seconds
IdleCycle Limit : 16 Cycles
Dynamic Counter : Enable
R/W Queue Bypass : 16x
Bypass Max : 07x
32 Byte Granularity : Disable

Might be different bios giving different results. I'm on 704-2BTA, yellow slots

thanx bro, i'm on 166 divider coz i dun have a higher multi...

Drive and data strength diff too in nf3u, dunno how to interprete 'em .

will try it out & update u guys here.

me too but, manage to fiXX it
my BE-5 chipset can works wif AUTO(most value) setting. suddenly, i load default setting my BIOS, then this BE-5 working juz fine. currently burnin/ stress this 2gb kit at 220Mhz x11, 1:1, 2-3-3-5. runnin' dual channel, 1T, LDT=4x, vcore=1.4, vdimm=2.6

ps :
pass 32mb calc.
pass 3dmark05
pass pcmark05
will update screenie soon

thanks bro babyelf 4 such nice RAms

This post has been edited by gengstapo: Dec 11 2005, 01:18 AM

This is wat i get, any higher gimme thousands of errors in memtest. Not usin booster thou, onboard vdimm limit to 3.0V (3.2V max in bios) since my +3.3V rail is really weak(3.13V). Hopefully things will get well after i put my booster on.



thanx gengstapo, I'll try out ya method too.

ah_khoo,

too bad that your psu limited the onboard vdimm. still haven't get the booster? i tot you got it now.

hurmm.. i found tat my BE-5 can max stable at 1:1 wif 220HTT(2-3-3-5)
towards this value will keep-on restarting/ BOOTing

at 5:6, highest can go currently at 260HTT(2-3-3-5)

still looking for best setting haihzz.. mebe took about this whole day to tweek

This post has been edited by gengstapo: Dec 11 2005, 11:26 AM

I got it ledi, but left it in jb last few days, 4got to bring it here...

it takes more than juz few days to get d best outa ram, mebe weeks for me...

This post has been edited by ah_khoo: Dec 11 2005, 11:39 AM

TEAM BH-5 owners. Anyone managed to get memtest free > 260MHz, Vdimm and what's your week i'll like to know....

hi, how high can your be-5 go with loose timings without the divider?

havent tried wif loose timming(& i dont like it)
wif 2-3-3-5 highest can get 280HTT x 9multi. downclock to ddr333
vdimm at 2.6 while vcore at 1.6

i oso tried last nite wif 1:1 setting but timming loosen 3-4-4-10, can boot at 280HTT x 10multi.. but aint stable. mebe need to be tweek summore

wat u think??

This post has been edited by gengstapo: Dec 11 2005, 03:35 PM

so, the be-5 behavior is more like bh-5/6 or tccd/uccc ?
[voltage, timing...]

for me its more alike to be TCCD/ TCC5 becoz higher HTT need loose timming but lower voltage

after addin some juice on 'em, mine give less errors in memtest.

currently on 235MHz, 3.2V BIOS set...

need some more tweakin i guess...

how bout u guys?

mine Team BH-5 highest memtest stable 258fsb 3.7V

I think it looks more like D5 instead of TCCD/TCC5

for BH5 mosfet cooling (beside jp17 jumper) is very important on the DFI..

read to write will only work up to 250 or so at 1.. after that you will need 2..

i'll be bringing one rig back and will probably meet up with someone and try his pair and see how it goes..

for the DRAM Drive Strength & DRAM Data Strength,

which is tighter? 8/3 or 7/1... i mean higher value is tight or loose?

there's no tighter afaik..

the higher the drive strength the more stability you get for an overclocked ram.. same goes for data strength

thanks for your info.

i'm new to dfi nf4 ultra-d, could you tell which setting is related to stability and which one is related to overclock(i mean like tcl,trfc,trc and etc.).

@Babyelf

Ur Team Xtreme TCCD 1GB how much HTT it can go?I mean the highest.

This post has been edited by simpleguy: Dec 27 2005, 02:30 AM

I haven't gotten the DDR500 one yet.. the previous TEAM TCCD i had goes up to ~ 600+ 2.5-3-3-7

Gskill 2x512MB GH, 250 @ 2-2-2-5 need 3.43v



alamak... now i only noticed that i'm stressing the CPU

This post has been edited by gsan: Dec 27 2005, 11:52 PM

From your screenie it shows that your rails are pretty high especially +12v. SmartGuardian showing 12.48 and i'm quite sure that it could be 12.7v / 12.8v. Should watch for that....

yeah SG normally reports it a little lower
wat psu are u using?

he is using Acbel i guess

thanks for your info. i'll check soon.

Acbel iPower 450W.

this time wont be wrong

Gskill 2x512MB GH, 250 @ 2-2-2-5 need 3.43v

mine can't do 233mhz....anybody can help me?
lots of error at then end of test 5....

agreed.... higher doesn't always good. u might hurt other components that use +12v rail... creating unecessary heat etc.