Why is'nt anybody trying buffalo micron G??

$80 a newegg. And unless they've screwed up again, it's labled as MG= micron G, yes the same stuff found on the ballistix you all are dying to get at twice the price.

Here's the link look Model# DD4333-512/MG

http://www.newegg.com/app/ViewProductDesc.asp?description=20-150-617&depa=1

Any comments before I order a couple and try?

Actually, in the original thread about the Buffalo Tech PC3700, one of the guys ordered the PC3200 from newegg and received the "G" revision chip, but was unable to do very much with it, either. Micron "tweaks" the chips used in the Ballistix to help them perform better, so they are not exactly the same as you would get on the ram from newegg.

Here is a link to that thread.

http://www.ocforums.com/showthread.php?t=311457

Reefa_Madness thank you for pointing that out/ One thing about buffalo thats cool, unlike other mem makers, thier model numbers seem to leave no question what chips they have on them

I got some xxx/WC here meaning winbond CH-5 and had some xxx../WB the good BH-5. I sold the later for some KHX.. Anyway wish me luck with /MG...cant beat the price .

The model numbers tell the story, and their website lists the various chips used as well, then they don't cover them up with heatspreaders.

You most definetly have to give them credit for that. They don't try to hide behind labels. Kingston ValueRam was that way, the product code also told which chips (HY for Hynix, MT for Micron Technology, etc) were being used, and they were visible through the window in the retail box. This is no longer the case, as they now use new labels that block the lettering on all of the chips which are visible through the window in the Retail packaging. I guess they caught on to us buying up all of the kvr with Winbond, Hynix, Samsung, Micron, etc. instead of the HyperX products.

Reefa_Madness said:

The model numbers tell the story, and their website lists the various chips used as well, then they don't cover them up with heatspreaders.

You most definetly have to give them credit for that. They don't try to hide behind labels. Kingston ValueRam was that way, the product code also told which chips (HY for Hynix, MT for Micron Technology, etc) were being used, and they were visible through the window in the retail box. This is no longer the case, as they now use new labels that block the lettering on all of the chips which are visible through the window in the Retail packaging. I guess they caught on to us buying up all of the kvr with Winbond, Hynix, Samsung, Micron, etc. instead of the HyperX products.

Click to expand...

yes i saw that last time i was in bb. that sucked. i also learned a real good lesson concerning good overclocking chips. i assumed that since all the kvr i bought with desirable chips ie: hynix/winbond/ micron really ran well, these same chips in other brand memory modules would overclok accordingly.i found out how wrong i was. (namely the buffalo pc3700 w/micron-5b c) In this case, speed rating (good testing)and chip selection is just as important.. i still find it hard to believe they tested their micron pc3700 anywhere near as well as they should have..

They seem to have done the same with their bt-d43 based DDR533. During a conversation that I had with one of their tech support guys, he told me that this ram should run at least 250. I'm thinking to myself, you have it rated as 266MHz capable ram and your saying it should run 250, heck yeah it should run 250...and MORE! When I asked about it, he gave me the "on certain motherboards" answer. It is not his fault, he doesn't rate them, this is done in Japan. Somebody over there is doing shoddy work or is very greedy and doesn't really care whether or not they hit rated speeds.

I should send them a link to the Adata threads so they can see what DDR533 is supposed to do.

btw - reached a milestone of 1,000 (mostly very long) posts today.

Ya Bigtoe/tony at xtreme said the same as you're saying flapperhead. He said they spend major dollars and time binning these chips. Thing is, what if all chips hynix sends out on a order does 260-275... then KVR and the like have no choice but to give us DDR 550 when that's the modules they got the best price on and chose to use in thier budget ram. Worth the risk for half dollars or less IMO.

Sure it's important, but how important depends on yeilds id think. Nothing guranteed unless you fork over $350

Reefa_Madness said:

They seem to have done the same with their bt-d43 based DDR533. During a conversation that I had with one of their tech support guys, he told me that this ram should run at least 250. I'm thinking to myself, you have it rated as 266MHz capable ram and your saying it should run 250, heck yeah it should run 250...and MORE! When I asked about it, he gave me the "on certain motherboards" answer. It is not his fault, he doesn't rate them, this is done in Japan. Somebody over there is doing shoddy work or is very greedy and doesn't really care whether or not they hit rated speeds.

I should send them a link to the Adata threads so they can see what DDR533 is supposed to do.

btw - reached a milestone of 1,000 (mostly very long) posts today.

Click to expand...

yep, it only takes a short time to lose a cutomers trust and a long time to get it back.. congrats, on your 1000th post ive really enjoyed the many discussions we have had in this forum..

I just tested the Buffalo 3200 'G' chips (512mb). 0418 mT 46v32M8
TG -5BG. It was about as extensive as you can do in 6 hours

At 2T these are 230mhz sticks on "3"s settings (3-3-3-X). This is at 3v because I mainly was at that, except for the max test. It's almost as if they tested these on Intel rigs because they are more like high fsb/high lat sticks instead of low fsb/low latency. Max at 4's and 3.3v is:

3.3v 3-4-4-11 2T



At 1T the best I could get was 2.5-2-2-6 @200FSB. AT 1T these are 200fsb sticks because I could not get 1 mhz over 200 or it would error in memtest86+.

2.5-2-2-6 200fsb@3v 1T


These are not the same as the Ballistix -5BG, but they are not bad either. They may go to 230mhz @2.8 and looser timings, but they will not go more unless you can give them 3v and over.

It also passed @210mhz on an A64 rig (K8T NEO no agp/pci lock) at 3-3-3-6 and 2.7volts. If I had time to try some 4's in the middle and 2.85v it would have gone higher.

As I am taking my time trying to match 2x256 HyperX CH5 for 1 Gig that top out @257mhz and has their sweet spot @246mhz where they pass 3Dmark2001, prime, ect. and 2-3-2-11 timings; I'll have to pass on this stick. I need a 512mb stick that will at least do that.

Wow thanks for the bredth of testing. As you say better than advertised but no diamond in the rough...no ballistix stick. What voltage did you use to run 2.5-2-2 1T? That's still very valuable for AMD 64.

I got a stick of the Crucial 8T stuff the other day. I think it has 5B-G chips. This is where I stopped:



That's with 2.75v. I know it's not as good as the Ballistix, but it was $85. I wanted to give it a run-through. I'm going to order another stick to hopefully pair it with. Hey, maybe I'll even get another two and run 1.5gb. Honestly, I'll probably end up running the RAM at 200mhz/2.5-2-2-6 simply because there isn't much difference between 200mhz/2.5-2-2-6 and 240mhz/3-3-3-8.

You should start a new thread Shade00, those some good results many builders would find compelling. Rock solid company like crucial with lifetime warranty and thier ram speaks for itself...

3v even, on the 2.5-2-2-6 1T. I had it there on most everything just because I didn't want to starve it trying to see what it will do. It really needs to be over 2.7 to do anything.

I can step it down on the vDimm till it borks, though I'll probably have to do a Cmos reset. Let me know if that would be useful info. I was thinking that if it wasn't going a few mhz past 200mhz at 2.5-2-2-6, 2.5-2-3-11, or 2.5-3-3-11, that it wouldn't go down in voltage much. Maybe it would though. All of the above was quite happy @200mhz,, no memtest errors.

Everything went OK between the 2 extremes above, i.e.; 3-3-3-11 was fine to 227mhz@3v, 3-4-4-11 was fine to 240mhz@3v, 1/10v for every couple mhz from there to 245mhz (all 2T).