OC Report: Team Xtreem 2x1GB PC6400 C4s (TXDD2048M800HC4DC)

3

3oh6

Weaksauce
Joined
Dec 23, 2005
Messages
91
I recently settled on this kit of memory after being dissapointed with a set of Promos based 6400 C4Ds. These sticks aren't anything special that will be breaking records but another solid pair of D9GMH based modules for an excellent price. I don't post much anymore here but since there is so much talk lately of 6400 C4s here, I thought this type of testing may help some folks out in making buying decisions...enjoy.

teamlogo.png
I think everyone here is more than familiar with Team Group and their line of Xtreem memory so I won't bother with much of an introduction as it isn't necessary. I will mention that Team Group does offer a lifetime warranty on their memory and is very forward with letting you know what ICs are planted on the PCBs of each stick which is highly appreciated by this user. The fact that heatspreaders are up to the users discretion is very nice as well.

The modules I will be testing today are a retail sample I purchased. They can be found at DirectCanada.com for about $320CDN and Tank Guys has this kit listed for $287USD for you south of my border crowd. These are not hand picked or even a review sample that might have tested well, just your plain jane run of the mill package off the shelf if you will. Here is a quick list of the specifications I have put together:

* Part # TXDD2048M800HC4DC
* ICs: Micron D9GMH
* Available in 1GB Dual Channel kits as well
* PC2-6400 800MHz 4-4-4-12
* 64x8 DDR2 FBGA Chips
* 240pin, Non-ECC, Un-buffered DDR2 SDRAM DIMM
* Aluminum Heat-spreader
* 6 Layers Ultra Low Noises Shielded PCB
* Retail Package
* Operating Voltage: 2.0V - 2.2V
* Lifetime Warranty
Click to expand...

Now, with the boring pleasentries out of the way, time to get to the good stuff...eye candy!

Package:
Click for full size...
​
I have already mentioned that the Team Group memory does not come with heatspreaders attached. They come in a seperate package so that the end user can decide whether or not they want them on. This makes identifying the ICs without voiding warranty as simple as reading the label. For this review, the heatspreaders will not be touching the modules. I normally take them off all of my RAM, so Team Group has saved me a step ;)


Package Contents:
Click for full size...
​
Naked sticks look better IMO than those with heatspreaders anyway so I am partial to the plain looking memory pr0n but I have to admit, the silver heatspreaders that come with the Team Group memory are very nice indeed. Not nice enough to wreck the cool factor of stripped sticks but damn fine looking if I say so myself. Those will likely end up as a badge on something if not in a small frame on the wall...instant art. Identifying the ICs isn't brain surgery with naked sticks and we can clearly see D9GMH...is there really anything else? Other than "fatbodies" and D9GKX of course.


Install & Testing:
Click for full size...
​
Installation is straightforward and the first image is of the setup for the AMD testing. The second image gives away what will be used in conjunction with the P5B-Dlx for this review. The third image above is how the setup looked for testing on the P5B-Dlx. Here is a complete break-down of the hardware on each setup:

Code: 
[b]AMD Setup:[/b]                                 [b]Intel Setup:[/b]
MB: DFI LanParty NF590 SLI-M2R/G           MB: Asus P5B-Dlx Wifi
CPU: AMD64 X2 4000+ (CCBYF 0620UPMW)       CPU: Intel C2D E6300 (L626A453)
CPU Cooling: Asetek WaterChill             CPU Cooling: Thermalright Ultra-120
PSU: OCZ GameXStream 700W                  PSU: Silverstone Zeus 560W
HD: Seagate SATAII 80GB 8MB NCQ            HD: Seagate SATAII 80GB 8MB NCQ
OS: Windows 2K3                            OS: Windows 2K3
                                           EXTRA: Hipro-Tech.com DDR Maximizer
Ambient Temperature: 22-23C
I have decided on a self imposed limit for the initial results on this memory at 2.6v. With the DFI board and the Maximizer, voltage is unlimited, but I need this memory to actually do some work before I go and kill it so for now...2.6v is all it will see. Time for some benches...

All results are simply dual 32M SuperPi stable...that is it, that is all. No Memtest or Prime testing is or will be done. Dual 32M SPi stable at the timings I use are pretty much going to be Prime stable with or without a bit of work on the secondaries. I like dual 32M as a stability test for memory so that is what I will show.

Also keep in mind that the voltage reading from SmartGaurdian on the AMD results reads the VDIMM about 0.11v low from what is reported in the BIOS. Enjoy...

3-3-3-X:
Intel:
Click link for full size screenshot...
DDR800 :: 3-3-3-8 :: 2.45v
DDR820 :: 3-3-3-8 :: 2.55v
DDR830 :: 3-3-3-8 :: 2.60v

AMD:
CAS 3 didn't take too kindly to my X2 4000+, perhaps I will see if my 3200+ will be a willing participant later on in life.


4-3-3-X:
AMD:
Click link for full size screenshot...
DDR800 :: 4-3-3-5 :: 2.25v
DDR840 :: 4-3-3-5 :: 2.45v


4-4-3-X:
Intel:
Click link for full size screenshot...
DDR850 :: 4-4-3-4 :: 2.00v
DDR890 :: 4-4-3-4 :: 2.10v
DDR950 :: 4-4-3-4 :: 2.25v
DDR990 :: 4-4-3-4 :: 2.40v
DDR1015 :: 4-4-3-4 :: 2.45v
DDR1040 :: 4-4-3-8 :: 2.60v

AMD:
Click link for full size screenshot...
DDR840 :: 4-4-3-5 :: 2.00v
DDR880 :: 4-4-3-5 :: 2.10v
DDR930 :: 4-4-3-5 :: 2.20v
DDR960 :: 4-4-3-5 :: 2.30v
DDR1040 :: 4-4-3-5 :: 2.47v
DDR1070 :: 4-4-3-5 :: 2.60v


5-4-4-X:
AMD:
Click link for full size screenshot...
DDR940 :: 5-4-4-5 :: 2.00v
DDR980 :: 5-4-4-5 :: 2.10v
DDR1030 :: 5-4-4-5 :: 2.20v
DDR1070 :: 5-4-4-5 :: 2.33v
DDR1100 :: 5-4-4-5 :: 2.42v
DDR1120 :: 5-4-4-5 :: 2.50v
DDR1140 :: 5-4-4-5 :: 2.60v


5-5-4-X:
Intel:
Click link for full size screenshot...
DDR1020 :: 5-5-4-5 :: 2.40v
DDR1040 :: 5-5-4-5 :: 2.47v
DDR1060 :: 5-5-4-5 :: 2.60v

AMD:
Click link for full size screenshot...
DDR1130 :: 5-5-4-5 :: 2.20v
DDR1180 :: 5-5-4-5 :: 2.30v
DDR1200 :: 5-5-4-5 :: 2.45v
DDR1220 :: 5-5-4-5 :: 2.52v
Past DDR1220 my CPU runs out of steam without chilled water but this CPU is so badly cold bugged that it doesn't really go much past 3100MHz without a balancing act. Needless to say, this memory won't really be pushed to its limits on an AMD board until I get a better CPU. Bring on the 65nm goodness :D


5-5-5-8:
Intel
Click link for full size screenshot...
DDR1020 :: 5-5-5-8 :: 2.00v
DDR1080 :: 5-5-5-8 :: 2.10v
DDR1120 :: 5-5-5-8 :: 2.20v
DDR1160 :: 5-5-5-8 :: 2.35v
DDR1200 :: 5-5-5-8 :: 2.50v
Finally with the 0804 BIOS and a phase change on my E6300, the 1:1 memory divider was able to clock this memory at CL5. You will notice that the DDR1120+ runs are on the 2:3 divider, I just couldn't get 560FSB+ to run dual 32M stable 1:1. That is where my stock P5B-Dlx seems to run out of steam.

So there you have it...my preliminary results from this memory kit and I am quite pleased. This memory ran very smoothly and had no problems booting in the P5B-Dlx from a freshly reset CMOS. I hope to add more results as I get them but for the most part, the bulk of what I will be posting is up. Perhaps some 680i action is in order as soon as my P5N32-E SLI comes in so look for that.

And here is a single 32M that I did with this kit of memory. This run is a bit more extreme and done with voltage that I would not recommend for anyone :D

32m_553x7_hwbot-1.gif
 
This is the RAM everyone should be buying right now :D

Team has great stuff we can all afford, & it all OCs to sky high :D
 
I'm still extremely new to overclocking and pc building in general, but does leaving the heatspreaders off do anything else for the ram besides "improving looks." Does it hurt the ram any (I guess obviously not since you were able to overclock these, so are the heatspreaders really just for looks?)

I was going to go with Corsair but it sounds like that won't be a good idea since I would like to get into overclocking on this first build of mine.
 
actually, heatspreaders on are supposed to improve looks. off, slightly better cooling with a fan on the ram ;)

3oh6: seems you had the same problem i had on the amd platform :(

610.3@5-5-4-8-2T-1M-2.29v.png


stupid cpu limits :mad:
 
(cf)Eclipse said:
stupid cpu limits :mad:
Click to expand...
i hear ya, although my sticks aren't quite as good as yours. they aren't even close to be honest but that may have something to do with the CPU being used.

those bloody 65nm parts coming soon better clock to 3400MHz so we can at least max out decent sticks of memory.
 
seriously! i'm really looking forward to it :D
if it doesn't clock as well as my cheapass sempron, i'm gonna be pissed

i did some rough fuzzy math a while back that made me laugh, ended up with somewhere around 4.8ghz as a high end overclock. :confused: :rolleyes: :D
 
hahaha, i would be quite pleased if your fuzzy math was even remotely close but something tells me that is why it is called "fuzzy math" :D

i would like to see DDR1300 on AM2 so if the average overclock is around 3200-3250 i would be pleased iwth the first incarnation of the 65nm parts. if not, then i will be hard pressed to hide my dis-pleasure.
 
If you still have them in the Intel system, i'd be interested in seeing what the max they can do Orthos blend stable is.

I know that's not something you extreme testers like doing, but SuperPI is a joke for stability comparably, which is why i am curious.
 
sardonic.n7 said:
If you still have them in the Intel system, i'd be interested in seeing what the max they can do Orthos blend stable is.

I know that's not something you extreme testers like doing, but SuperPI is a joke for stability comparably, which is why i am curious.
Click to expand...
no problem, i will add it to my work board. right now they are priming on the AM2 setup but i should be able to get them back on the P5B-Dlx for Priming over night. i am brutalizing some 5400ULs on the P5B-Dlx rig right now :D

generally speaking, i have found dual 32M stable to be pretty close to Prime Blend stable, usually with loosening of secondaries i can stabalize them at the same frequency/timings/voltage. which max would you like to see Prime Blend stable...4-4-4-X or 5-5-5-X? also, was there a specific voltage you would like me to stay under? for Prime testing i will generally stay under 2.5v so how does 2.4v sound?
 
3oh6 said:
generally speaking, i have found dual 32M stable to be pretty close to Prime Blend stable, usually with loosening of secondaries i can stabalize them at the same frequency/timings/voltage.
Click to expand...
likewise, and trust me when i say i did a lot of research into the matter :D

sometimes, 32m will fail a few mhz BEFORE prime/orthos blend does ;)
 
<2.4V is fine.

Max 5-5-5 would be mainly what i'm interested in, unless it's barely any different than 4-4-4.

I'm thinking about what i mentioned about SuperPI, & i guess maybe dual 32M isn't so bad for testing RAM.

But it's terrible for testing CPU, that i can attest to.

I've had dual SuperPI 32M pass, yet had games crashing & locking up at the same speeds.
 
sardonic.n7 said:
I'm thinking about what i mentioned about SuperPI, & i guess maybe dual 32M isn't so bad for testing RAM.

But it's terrible for testing CPU, that i can attest to.
Click to expand...
dual 32M SPi is definitely a good way for testing memory, that is why i stick with it for my stability testing with memory when they first come in. it generally provides the fastest and best way to gauge a set of memory. your right though, it isn't good for testing CPU stability. with a CPU overclocking report i wouldn't be showing dual 32M SPI screen shots...that is where prime or running Rosetta is king.

i ran the max 4-4-3-4 overclock @ 2.40v and with the exact same settings as the dual 32M posted in the report, it errored out just before an hour. i then set the secondary timings to defaults and it ran for a good 6 hours:

4434prime-1.png


32m stable: 3-25-2-10-10-6-8-6-3
prime stable: 3-42-2-10-10-10-8-10-3

i am assuming tRFC was the only one that needed to be loosened up for added prime stability but i wanted to make sure it would run for at least the 6 hours i slept. even a tRFC of 35 might have been enough but there is very little performance loss going from 25-42, except for SPi benching.

i will check 5-5-5-8 at 2.40v tonight for you.
 
just bought this ram for my aw9d-max...thanks for the benchmarks :)
 
You must log in or register to reply here.
Back
Top