First Look: Teamgroup 2x512MB PC4000 BH-5

[s7e9h3n]

So you guys (and gals) say you want some fast sticks for benching? These modules right here may just be the answer. Slated for DDR500 speeds, 2-2-2-5 @ 3.3v (sorry I was wrong about the original 3.2v estimate), the potential for high OC's don't seem so out of reach any more. The sticks you see right here are the first set of PC4000 from Team's "Xtreem BH-5" line to be released to the public and I've got the honor of testing/torturing them first I've also been assured that the performance of the retail release of these kits will be very comparable - if not better - than this particular set. So no cherry-picking here Here's a few pics:

The heatspreaders on these sticks don't seem as suffocating as the standard ones we usually see on the RAM from Kingston, Mushkin, etc. I'd say that this is probably a better design in that sense. But personally, I've achieved my best results when the sticks are naked....sooooooo....... Note: Be advised that removing the heatspreaders from most modules of memory will void your warranty. Teamgroup's memory should not be an exception. If you end up borking your sticks and are unable to RMA due to the removal of the spreaders, please don't point to this thread and say "Well s7e9h3n did it".....


Yes, they ARE in fact rated @ PC4000, 2-2-2-5. Don't be fooled by the PC3200 sticker :


A closer look at the IC's and PCB. These sticks use 0551 BH-5 IC's on Brainpower 808 pcb's:


A full set of tests to come..........

[Lastviking]

Same pcb on my pc3200(brainpower) but my sticks are 0548T, i think xtreem test all sticks and take out the best and rate them pc4000.

[rookiekiller]

hm, lecker =o)
i want it 2

1 stick of my redline was killed after prime =o( now it has bootproblem, i have to boot with 1 stick then shutdown and put the second killed stick in mobo then it can boot and run stable like normal stable =o)


show us ur best result of these new power

[pl4y3r]

How much did u pay 4 this? in my country TeamGroup 2x512 3200 BH5 costs 192$
now show us some tests ;D
http://img526.imageshack.us/img526/5...koscbig2sw.jpg

[metro.cl]

nice where to get them??

[Lastviking]

How much did u pay 4 this? in my country TeamGroup 2x512 3200 BH5 costs 192$

About 200$ for the pc3200 they are not cheap..and the price has going up to know.

[dinos22]

looking forward to some test results.......BH5 PC4000 RAM hey.....this is going to be good

[bachus_anonym]

Let's go, Steven... Don't leave them all hanging here and make sure to bring some 32M with ya

nice where to get them??

By googling, I found one e-tailer here in US, that carries Team Group's UTT-BH5 kits (PC3200 for $135 and PC4000 for $177) - Computer Giants. No idea about their reputation, though...

[s7e9h3n]

Let's go, Steven... Don't leave them all hanging here and make sure to bring some 32M with ya
By googling, I found one e-tailer here in US, that carries Team Group's UTT-BH5 kits (PC3200 for $135 and PC4000 for $177) - Computer Giants. No idea about their reputation, though...

NP....just waiting so I can plug in my autocascade, but this load of laundry has to dry first...lol
AFAIK, this memory and for that matter ALL TEAM memory hasn't been officially released in the U.S. When it does, there should be TWO major e-tailers carrying the brand and computergiants isn't one of them

[Cpt Twitchy]

My predictions

ddr500 3.2v
ddr560 but I'm not sure on volts. Maybe 3.7v
ddr574 max bench stable at hmmm 3.85v

BTW YGPM

Tim

EDIT they don't have a very good rating at reseller ratings http://www.resellerratings.com/selle...seller_id=2891

[gigi77]

go up

[Coroner Kyle]

Is your laundry dry yet?

I really want some BH-5.

[s7e9h3n]

Is your laundry dry yet?

I really want some BH-5.

Hehehe...it's been dry for a few hours....I had to take care of a few other things, but I'll post a few results soon. I just need to wait till this 32m finishes.....

[s7e9h3n]

OK sorry for the delay, but I had to re-install windows because I had corrupted my OS when I was trying to set up my bench drive for 3dMark Here are my initial results from the Team 2x512MB PC4000 BH5....

First the Vital Stats:
TEAM XTREEM 2x512 DUAL CHANNEL PC4000 BH5
CABCE 0512MPMW FX57
DFI Ultra-D Rev. AD0
PC Power and Cooling 510 SLI
Autocascade cooled

All Vmem voltages listed represent what was set in bios and not read through any monitoring IC's......

DDR500, 3.2Vmem, Spi1m


DDR500, 3.2Vmem, Spi8m


DDR500, 3.2Vmem, Spi32m

[s7e9h3n]

Next up....255Mhz - No prob for these sticks

DDR510, 3.2Vmem, Spi1m


DDR510, 3.2Vmem, Spi8m


DDR510, 3.2Vmem, Spi32m

[Lastviking]

lol i see a funy thing the sticks are made before the ic on the sticks was made. If you look in i cpuz SPD.

[s7e9h3n]

Now this is where I spent a majority of my night tweaking. I went literally every tweak I knew of to try to get a bench done at 260mhz @ 3.2Vmem, but absolutely nothing came out of it I was, however, able to complete the set of 260Mhz runs at a lower Vmem than I have ever seen done at this memory speed:

DDR520, 3.3Vmem, Spi1m


DDR520, 3.3Vmem, Spi8m


DDR520, 3.3Vmem, Spi32m


That's all I've got for tonight. On a side note: I've run my autocascade for about 5 hours straight and it's not even hiccupped. Notice the increased vcores have absolutely no affect on the core temps of the cpu - which is a great testament to how much capacity this unit really has......I'll push these sticks some more tomorrow and try to get some 3d runs in with them as well (and hopefully at the same time take down the x850xtpe top mark in 3d01 )

[s7e9h3n]

lol i see a funy thing the sticks are made before the ic on the sticks was made. If you look in i cpuz SPD.

I think that the spd on the Kingston Bh5's were the same way, right?

[Lastviking]

One more thing i hope you test them with 3.4,3.5,3.6v also.

[Viss]

Same pcb on my pc3200(brainpower) but my sticks are 0548T, i think xtreem test all sticks and take out the best and rate them pc4000.

Thats me thought. Got a pair of PC3200 that needs at least 3.33v to finish 1M at 250, for 260 it needs 3.60v. Mine are 0601. If they started binning up for PC4000 in week 51 already i guess Giorgioprimo got lucky with his week 52 PC3200 sticks that do up to 28x...

[s7e9h3n]

One more thing i hope you test them with 3.4,3.5,3.6v also.

No need to hope, these sticks are going to get the full 3.6+V treatment as soon as I figure out their timings.

Thats me thought. Got a pair of PC3200 that needs at least 3.33v to finish 1M at 250, for 260 it needs 3.60v. Mine are 0601. If they started binning up for PC4000 in week 51 already i guess Giorgioprimo got lucky with his week 52 PC3200 sticks that do up to 28x...

No offense to Giorgioprimo, but in his tests, his sticks were running relatively loose timings. You really can't compare the results. He did, however, tighten his timings @ 272Mhz IIRC but it took 3.91V to get that stable. If these sticks continue to scale like they are, I'm hoping to be able to reach 270Mhz benchable @ ~3.55V Max. I guess we'll just have to wait and see....Also note that there is a subtle difference in the part numbers from Giorgioprimo, LastViking, and your sticks of the PC3200 flavor. I don't know if it was done on purpose or not, but they're definately not the same as these sticks

[gigi77]

it'a hand picked IC!

This Reviews is for DEMO, capability this IC

average i Think 270mhz 3.6v 2.2.2.5

[Viss]

No offense to Giorgioprimo, but in his tests, his sticks were running relatively loose timings. You really can't compare the results.

Agree but mine arent that tight either.

Also note that there is a subtle difference in the part numbers from Giorgioprimo, LastViking, and your sticks of the PC3200 flavor. I don't know if it was done on purpose or not, but they're definately not the same as these sticks

What differene are you reffering to?

yours: 200500115880
giorgio: 200500115975
mine: 200500116019

[s7e9h3n]

it'a hand picked IC!

This Reviews is for DEMO, capability this IC

average i Think 270mhz 3.6v 2.2.2.5

Hmmm, I think that all the PC4000 BH5 is handpicked though. Otherwise there wouldn't be any logical reason why there are known PC3200 sticks that have production dates both before and after these sticks. That would suggest (as LV and loco both have stated already) that TEAM is binning these PC4000 sticks from the current PC3200 stock.......

[s7e9h3n]

What differene are you reffering to?

yours: 200500115880
giorgio: 200500115975
mine: 200500116019

Hehehe...not the Serial Number.....the P/N or Part Number.....hint: It's on the other sticker BTW, your timings definately can't be considered as "loose." They look pretty good to me